Frog Story Slot Machine Free Online Casino Game by EGT

Frog Story

This EGT video slot game blends traditional elements of famous fairy tales with some surprises to keep players entertained and eager for a few more spins.

Check out our Frog Story review to find out what the gameplay and atmosphere are like. You can play this slot for real money or for free credits in our collection of free slot machines.

Once Upon A Time…

It's clear that Frog Story is set in a fairy tale world, and the designers have striven to create an atmosphere of fantasy and mystery.

The action is set in a forest, where lush vegetation and intricate tree roots merge in the middle of the path. You can catch glimpses of clear streams bordered by flowers and bathed in soft sunlight. As you can see, it's a very idyllic and peaceful place. I'd love to explore this enchanted forest a little more.

As usual with EGT titles, there is no background music, but instead, there are some sound effects that play during the spinning of the reels and after each win.

Venturing Further Into The Forest

When it comes to gameplay, you can't go wrong with Frog Story. There are 5 reels and 20 fixed paylines on which you can bet, and some paylines cannot be disabled. Therefore, there is only one parameter that can be changed - the size of the bet. Choose your bet: 20, 40, 100, 200 or 400 coins using the shortcuts at the bottom of the game screen. Once you have made your selection, the reels will start spinning automatically.

After each win consisting of an identical combination of paylines, the corresponding button appears, and you can gamble new coins by clicking on it. The mini-game is a simple game of guessing the color of the hidden card. If you get it right, you will get double coins, but if you get it wrong even once, the game is over and your coins are gone forever.

Another option is to use the orange auto play button on the left side of the screen to rotate the reel with as much as you want. Let's sit down and see the winning piles! Keep in mind that each victory is a chance to trigger a jackpot card bonus game, the only entrance to the FROG STORY progressive jackpot. There are four jackpot cards, displayed at the counter at the top of the screen. Once the game starts, choose a back card until you have three same suits. You can get a huge jackpot that corresponds immediately.

Another original point of game play is the topping reel function. After each victory, the related symbols disappear, and a new symbol comes down from above. If a new combination of wins appears, its value will be doubled. Each time a new combo is born with the topping reel function, the magnification increases by one.

Magic Is In The Air

Let's take a look at the FROG STORY payout. Starting with several classic symbols: Jack, Queen, King, and Ace drawn with colorful calligraphy. Although these symbols are not so many dividends, the topping reel function can open a better victory.

Next, mysterious treasure chests, magical medicine, princess and prince, and finally ol d-fashioned wise wizards appear. Each symbol is full of details and starts to move when the winning combination is established, so the games are very vibrant and rich in entertainment.

However, FROG STORY has more magic ...

Final Spell

There are two more symbols in the pneito. The first symbol is the scatter symbol castle. Regardless of the position on the reel, you can earn tremendous prizes: If you have five scatters, you can earn 500 times the bet.

Finally, a famous frog appears as a wildcard for this game. Therefore, you can substitute any symbols other than the scatter card. Also, when it appears in the 2nd, 3rd, and 4th reels, it will automatically expand to the entire reel, inducing further extra combination!

A Magical Game For All

FROG STORY combines classic slot game elements with innovative elements such as topping reel function and i n-game animation. This makes it a lively and fun game loved by many players.

However, the setting limit, especially the number of fixed paylines, will be exceptional for some players.

Frog Story Slot FAQs

Can I play the FROG STORY slot without charge?

If you are a casual gamer, you can enjoy FROG STORY slot machines for free on Vegasslotsonline. We provide thousands of online slots demo versions.

Can I get real money in the FROG STORY online slot?

You can play FROG STORY slot machines with real money. First, deposit some cash, bet on the bet, and try your luck on the reel.

Can I play a FROG STORY slot machine in my country?

The rules and regulations on online gambling vary from country to country, so please check the casino from a country where you can play the FROG STORY slot.

Why do I need to play a FROG STORY slot?

If you like fantasy and thrill, the FROG STORY slot is perfect. This slot is one of the newly added Selection of EGT slot games.

Free to Play EGT Slot Machine Games

10 Burning heart
10 Glossy hot
10 Power hot
100 Valquee Fruit
100 burning hot
100 cats
100 golden coins
100 power hot
100 super hot
2 dragon
20 Bulky fruit
20 Burning hot
20 Dazling hot
20 diamonds
20 Extra Crown
20 Glossy Hot
20 Golden Coins
20 Hot Blast
20 Joker dice
20 Joker reels
20 Power hot
20 Super hot
20 Super Hot Egypt Quest
20 Wild Goal
27 dice
27 Eternal Hot
27 victory
30 spicy fruits
30 Summer bliss
40 Almighty Ramses II
40 Balky fruits
40 giant fruits 6 reels
40 Burning hot
40 burning hot 6 reels
40 Golden Coin
40 hot & cash
40 king
40 Lucky King
40 luxury dice
40 mega clover
40 power hot
40 super hot
40 Ultra Spin
5 Burning heart
5 Dazling hot
5 Glossy dice
5 Glossy Hot
5 Great Star
5 Juggle Fruit
50 Battle of Amazones
50 horses
7 & amp; Crystal
7 & amp; Crystal dice
81 wins
Action money
Age of Troy
Almighty Ramses II
Aloha party
Amazing Amazonia
Battle of Amazones
Amazones story
Ancient dynasty
Golden art
Aztec forest
The glory of Aztec
Blue heart
Blue sea
Book of Magic
Brave Cat
Valquee fruit
Burning hot
Burning hot 6 reels
Burning hot Egyptian quest
Candy palace
Caramel hot
Casino mania
Cats Royal
Dinosaur and dinosaur
Circus Brilliant
Circus Brilliant Egyptian Quest
Cocktail dice
Cocktail rush
Coingobler
Coral island
Crazy bugs II
Dancing dead
Dark Queen
Diamond Plus
Diamond Plus
Dice high
Dorothy Fairy Country
Dragon hot
Dragon Reborn
Dragon reel
Dragon spirit
Easter Frog
Egyptian Sky
Egypt Sky Egyptian Quest
Emperor's palace
Empress charm
Extra Crown
Extra Crown Classic
Extra Joker
Extra Stars
Extreme Hot
Fast money
Burning hot
Flaming Hot 6 reel
Framing Hot Extreme
Forest band
Forest story
Fortune Spell
Frog story
Fruit & gold
Fruit kingdom
Fruity time
Lucky game
Leonardo genius
Gold dust
Cleopatra's grace
Great 27
Great adventure
Great Empire
Good luck in Greece
Halloween
Halloween Egyptian Quest
Hot & cash
Hot deco
Ice valley
Imperial Wars
Incagold II
Jungle adventure
Kamakura
Kangarouland
Kashmir Gold
Kemet's treasure
Knight's heart
Legendary Rome
Like a diamond
Lucky & wild
Lucky buzz
Lucky Hot
Lucky wood
Magellan
Magic Guardians
Majestic Forest
Maya spirit
Mighty Sparta
More Like a Diamond
More lucky & wild
Mythology treasure
Ocean rush
Oil Company II
The glory of Olympus
Orient story
Orient Story Deluxe
Penguin style
Phoenix Star
Pinup Queen
Power hot
Queen of Rio
Rainbow Rack
Rainbook Eien
Retro cabaret
Retro style
Rich World
Richness factory
Rise of La
Rise of La: Egyptian Quest
Route of Mexico
Royal Garden
The secret of the royal family
Savanna life
Alchemy secret
Shining Crown
Spanish passion
Stoichikov Episode 8
Summer bliss
Super 20
Supreme Hot
Sweet cheese
Big Journey
The Explorers
Great Egypt
London secret
Alexander story
Alexander Story 2
White wolf
Princess Oya Yubi's Dream
Tyrol Fest
Ultimate Hot
Vampire Night
Venice Doro
Versailles Gold
Viking rising
Volcano Wealth
Wild traces
The charm of a witch
Wonder tree
Wonder Heart
Zodiac Wheel

EGT Slot Machine Reviews (No Free Games)

20 Super Hot Skarabe
40 Super Hot Skarabe
5 Dazling hot scalabe
Age of Trois Carab
Amazones Battle Scalabe
Blue Heart Skarabe
Dice book
Burning hot scalabe
Captain Nemo's trip
Caramelhot Skarabe
Clover Islands
Dice & Roll Scarab
Extra Stars
Extreme hot scalabe
Flame hot skirab
Lucky Hot Skarabe
Oil Company II Scalabe
Rise of Razkarabe
Royal Secret Scarab
Sharwood's secret
Shining Crown Scarab
Sprem hot scalabe
Ultimate Hot Scalabe
Zodiac Wheels Calabe

Casino Bonuses

See the bonuses obtained from the top online casinos, from welcome packages to reload bonuses.

Online Gambling

Are you ready to learn all about online gambling? See the 2024 best gambling site ranking.

Mobile Slots

No matter which device you play, you can enjoy your favorite slot on mobile. Please see how to get started.

Live Dealer Casinos

Live dealer games are the most innovative and immersive games of online casinos. Let's learn about the top live deeply casino with our expert guide!

Our Favourite Casinos

Claim Free Spins, Free Chips and Much More!

By signing up, you agree to receive bonus offers and other emails from VSO, confirm that you are of legal gambling age in your country of residence, and that you have read and accepted the Privacy Policy. You can unsubscribe at any time. This site is protected by reCAPTCHA and is subject to the Google Privacy Policy and Terms of Use.

Quality of Service-Aware Multi-Objective Enhanced Differential Evolution Optimization for Time Slotted Channel Hopping Scheduling in Heterogeneous Internet of Things Sensor Networks

The emergence of the Internet of Things (IoT) has attracted significant attention in industrial environments. Such applications require stringent latency and reliability criteria to be met. To address this, the IEEE 802. 15. 4e standard introduces a new medium access control (MAC) protocol called time-slotted channel hopping (TSCH). Designing a centralized scheduling system that simultaneously achieves the required quality of service (QoS) is challenging due to the multi-objective optimization nature of the problem. In this paper, we introduce a new optimization algorithm, QoS-aware multi-objective enhanced differential evolution optimization (QMDE), which is designed to address QoS metrics such as latency and packet loss for multiple services in heterogeneous networks while also achieving the expected service throughput. Through co-simulation of TSCH-SIM and Matlab, R2023a, we conducted multiple simulations with diverse sensor network topologies and industrial QoS scenarios. The evaluation results show that the optimal schedule generated by QMDE can effectively meet the QoS requirements of industrial services of closed-loop supervisory control and condition monitoring in sensor networks from 16 to 100 nodes. Through extensive simulations and comparative evaluation with TASA (Traffic-Aware Scheduling Algorithm), this study reveals the superior performance of QMDE, achieving significant improvements in both PDR (Packet Delivery Ratio) and delay metrics.

Keywords

1. Introduction

In the context of Industry 4. 0, the fusion of advanced technology, especially the integration of the Internet (IoT), has revolutionized industrial processes. This paradigm shift focuses on how smart devices can work and to enable seamless communication and data throughout the manufacturing ecosystem. In order to achieve an industrial 4. 0 vision, the design of the wireless sensor network will be greatly changed, and the number of sensors monitoring and reporting the sensed data will be increased, and the complicated mechanical communication is supported under severe QOS (Quality of Service) requirements. You need to. The topology of the wide band wireless sensor network (WSN) needs to cover all the sensors of the entire environment and enable data transmission to the central sync node. The node is aggregated and treated information by automated and optimized routines, streamlining industrial operations.

In the IoT network area, one of the important factors that contributes to the connection of the industrial wireless network is the TSCH (Time Slotted Channel Hopping) mechanism derived from the IEEE 802. 15. 4 revision [2, 3].

Efficient scheduling in industrial wireless networks is extremely important to optimize network performance and resource use. The wel l-designed schedule is a timely data transmission, minimizing collision, waiting time, and packet loss. In particular, the industrial sensor network faces the task of accommodating sensors with different packet rates and QOS requirements, making it difficult to determine the optimal transmission schedule. In the Industry 4. 0 context, the fusion of advanced technology, especially the integration of the Internet (IoT) of things, revolutionized industrial processes. This paradigm shift focuses on how smart devices can work and to enable seamless communication and data throughout the manufacturing ecosystem. In order to achieve an industrial 4. 0 vision, the design of the wireless sensor network will be greatly changed, and the number of sensors monitoring and reporting the sensed data will be increased, and the complicated mechanical communication is supported under severe QOS (Quality of Service) requirements. You need to. The topology of the wide band wireless sensor network (WSN) needs to cover all the sensors of the entire environment and enable data transmission to the central sync node. The node is aggregated and treated information by automated and optimized routines, streamlining industrial operations.

Efficient scheduling in industrial wireless networks is extremely important to optimize network performance and resource use. The wel l-designed schedule is a timely data transmission, minimizing collision, waiting time, and packet loss. In particular, the industrial sensor network faces the task of accommodating sensors with different packet rates and QOS requirements, making it difficult to determine the optimal transmission schedule. In the context of Industry 4. 0, the fusion of advanced technology, especially the integration of the Internet (IoT), has revolutionized industrial processes. This paradigm shift focuses on how smart devices can work and to enable seamless communication and data throughout the manufacturing ecosystem. In order to achieve an industrial 4. 0 vision, the design of the wireless sensor network will be greatly changed, and the number of sensors monitoring and reporting the sensed data will be increased, and the complicated mechanical communication is supported under severe QOS (Quality of Service) requirements. You need to. The topology of the wide band wireless sensor network (WSN) needs to cover all the sensors of the entire environment and enable data transmission to the central sync node. The node is aggregated and treated information by automated and optimized routines, streamlining industrial operations.

Existing studies focus on optimizing a specific metric within the scope of a single application, often overlooking the QoS requirements specific to each application and the common scenario of multiple applications operating simultaneously on the same network. Each application needs to meet its own QoS demands, rather than simply minimizing or maximizing a metric. Furthermore, many studies assume a fixed packet rate, ignoring the heterogeneity of packet rates typical of industrial environments. The objective of this paper is to develop a TSCH scheduling algorithm that can accommodate these diverse QoS demands across multiple applications in wireless sensor networks. To the best of our knowledge, this approach is the first attempt to simultaneously consider heterogeneous application specifications, especially packet rates, and the challenge of meeting the QoS demands of multiple simultaneous applications.

In this paper, two main QoS requirements are considered: delay and PDR. Furthermore, multiple flows are specified, each responsible for the execution of an application with its own QoS requirements. This paper deals with optimizing multiple objectives simultaneously. The QoS optimization problem is a combinatorial optimization problem and is NP-hard [4, 5]. Because this multi-objective optimization problem is complex, we developed a QoS-aware multi-objective extended differential evolution optimization (QMDE) method. This method facilitates the determination of an optimal TSCH schedule that satisfies the QoS requirements, especially delay and packet loss (important parameters for each application in industrial environments), and simultaneously satisfies the required throughput of a centralized heterogeneous sensor network.

Achieving both delays and packet loss requirements over multiple data flows in the WSN constituting a multipurpose optimization problem. This is because it is a clear indicator with a goal that delays and packet loss may conflict. Reducing delays can increase the risk of packet loss due to several factors. For example, the sensor may send data more frequently to reduce delays. Such an increase in data will saturate the network and cause congestion. In congested networks, if the buffer is full or the packet exceeds the allowable sending deadline, a particularly sensitive application may drop packets.

In 6, 7], the purpose was to achieve the desirable throughput and at the same time identify the minimum slot frame. The minimized slot frame size corresponds to the reduction in delay [8], but depending on the application, it may allow a slight delay to prioritize other important QOS indicators, such as minimizing packet loss. be.

2. Related Works

In this study, we will introduce a new multipurpose extension evolution optimization approach aimed at satisfying various QOS requirements according to the introduced applications. This method is a pioneer that is not adopted in existing literature. The main contributions of this study are summarized in the following:

Industrial radio sensor network In order to deal with the issues of satisfying the QOS demands of various applications in the environment, in terms of time complexity of the customized de (PCDE) optimized algorithm announced in [7] It has been strengthened, shifted to a multipurpose, and created a schedule without conflict.

We have a PCDe algorithm to screw the node only when the node holds a packet (multiple) in a queue to avoid unnecessary transmission scheduling, which is an inefficient use of TSCH cells. It has been extended.

The configuration of this article is as follows. First, review related research in section 2. Next, the problem definition is explained in section 3. QMDE's methodology is described in detail in section 4. In section 5, the performance is evaluated, and the simulation setup and evaluation results of the proposal method are explained. The limits of the algorithm are described in section 6. Finally, section 7 describes the conclusions of this paper and the direction of future research.

No n-competing TSCH scheduling algorithms proposed in 9] aim to minimize transmission delays by shortening the slot frame length. The authors introduce the concept of "waves", and each node sends data at least once during this period. With this approach, nodes close to sinks may experience high traffic and queue overflows, which leads to an increase in delays. Furthermore, the sensor sends a fixed number of packets without considering the heterogeneity.

Similarly, the reference [10, 11, 12, 13, 14] ignores the heterogeneity of the network and focuses only on the fixed packet rate. Furthermore, in [15, 16], each node is assigned to a single cell, so that the performance of the algorithm may be reduced when the traffic load is high.

Traffic-Aware Schedularing Algorithm (TASA) [17] uses the principle of matching and coloring to determine a pair of transmission and reception nodes without collisions or interference. This approach follows a structured approach. First, use the concept of matching from the graph theory to identify a pair without a collision of transmission and receiving nodes based on the integrated queue size. After that, these selected pairs are assigned to different channels while guarantee interference avoidance by coloring mechanism. TASA has no mechanism to prioritize sending based on the required QOS requirements. In other words, a priority cannot be given to certain types of traffic that require a guaranteed service level. In addition, TASA works effectively with very low packet rates, such as one packet per hour, but is quite complicated when handling a high packet ratio. This complexity can cause problems in practical implementation, especially in an environment with a wide range of traffic requests.

The orchestra scheduling technique incorporated in the IEEE802. 15. 4 standard provides dynamic and effective adjustments for wireless communication in TSC H-based networks. Despite its advantages, Orchestra has a limit, especially when the traffic load is high. In such a situation, Orchestra scheduling can cause congestion delays and packet loss, ultimately affect network performance and have difficulty coordination.

DEAC's research [19] aims to improve Orchestra's receip t-based scheduling policy and support high traffic load. This includes adjusting the static schedule when the latest child node connected to the gateway experiences congestion. If the child node buffer exceeds the excess values, the child node will notify the gateway, and the gateway will use the hash function to increase the receiving time slot. However, since this method tackles collision issues only on gateway, other nodes of the entire network may remain unresolved, leading to scalability issues.

Several studies, including 20, 21, 22, 23], have proposed adaptation strategies to manage data traffic. These approaches propose a method of dynamically adding or deleting cells based on the fluctuating traffic status.

H. H. NGUYEN-DUY and others [24] proposed a scheduling algorithm based on strengthening learning (RL) using Q-Learning for schedule design. Q-Learning is a widely used RL method, especially for small optimization issues. However, in the heterogeneous IoT network, the state space is significantly large, and the table method such as Q-Learning becomes inefficient. [25]

GYAWALI et al. [26] applied deep enhanced learning (DRL) to the concentrated channel allocation in the vehicle network, but this approach is not very effective in dynamic IoT networks, which are difficult to obtain a complete channel status information. YE et al. [27] deals with this with a distributed DRL approach, and uses the Actor Clicky method to balance the solution by combining intensive value functions and distributed policies. DQ N-based algorithms are also used in industrial environments, but are suitable for rea l-time and dynamic scenarios, while our focus is the optimization of TSCH scheduling in a static or sem i-quiet environment.

3. Problem Definition

Various methods are proposed for TSCH scheduling [14, 28]. These TSCH scheduling algorizms have multiple applications executed simultaneously in the industrial wireless sensor network, but delayed, energy consumption, packet delivery, etc. in a single application context. It has been developed mainly to optimize specific metrics.

In the industrial environment, the specifications are different for each application, and QOS requirements vary. As a result of this scenario, a variety of data traffic must be generated, and each must have a demand for delays and reliability, and to be satisfied to guarantee the appropriate functions of the application. The task is to devise a TSCH schedule that synchronizes packets generated between various applications while satisfying each QOS standard. In order to deal with these requirements, not only minimizes slot frame sizes and delays, but also filling the required QOS parameters. In contrast to pr e-research, which focuses only on minimizing delay [15, 19] and maximizing reliability [11, 29, 30], this paper has multiple specified wireless sensor network. Treats the task of satisfying the QOS requirements.

Another limit of the current TSCH scheduling algorithm is the premise of a fixed packet rate, and does not take into account the uniformity of packet rates, which are often characterized by rea l-world applications. Traditional research has often overlooked the impact of heterogeneous packet rates in the industrial sensor network, as explained in [9, 10, 11, 12, 13]. These studies generally focus on fixed packet rates, simplifying the subsequent transmission of each sensor node. These approaches may not be enough in scenarios, where packet rates change significantly, do not take into account the complexity caused by traffic changes in the actual industrial environment.

Some solutions are dealing with traffic diversity [15, 17, 19, 31], but it is still important to develop QOS solutions that are optimized for heterogeneous wireless networks. It is. As far as we know, the existing approach does not fully consider scenarios where multiple applications with different QOS requirements and packet rates are executed on a single network. This oversight has revealed the substantial gap in the current research content, as optimizing industrial sensor network is indispensable for various applications with different performance requirements. In the industrial environment, the specifications are different for each application, and QOS requirements vary. As a result of this scenario, a variety of data traffic must be generated, and each must have a demand for delays and reliability, and to be satisfied to guarantee the appropriate functions of the application. The task is to devise a TSCH schedule that synchronizes packets generated between various applications while satisfying each QOS standard. In order to deal with these requirements, not only minimizes slot frame sizes and delays, but also filling the required QOS parameters. In contrast to pr e-research, which focuses only on minimizing delay [15, 19] and maximizing reliability [11, 29, 30], this paper has multiple specified wireless sensor network. Treats the task of satisfying the QOS requirements.< S 2 , S 3 , S 4 , S 5 , S 6 >Another limit of the current TSCH scheduling algorithm is the premise of a fixed packet rate, and does not take into account the uniformity of packet rates, which are often characterized by rea l-world applications. Traditional research has often overlooked the impact of a heterogeneous packet rate in the industrial sensor network, as is an overview in [9, 10, 11, 12, 13]. These studies generally focus on fixed packet rates, simplifying the subsequent transmission of each sensor node. These approaches may not be enough in scenarios, where packet rates change significantly, do not take into account the complexity caused by traffic changes in the actual industrial environment.< S 2 , S 4 >Some solutions are dealing with traffic diversity [15, 17, 19, 31], but it is still important to develop QOS solutions that are optimized for heterogeneous wireless networks. It is. As far as we know, the existing approach does not fully consider scenarios where multiple applications with different QOS requirements and packet rates are executed on a single network. This oversight has revealed the substantial gap in the current research content, as optimizing industrial sensor network is indispensable for various applications with different performance requirements. In the industrial environment, the specifications are different for each application, and QOS requirements vary. As a result of this scenario, a variety of data traffic must be generated, and each must have a demand for delays and reliability, and to be satisfied to guarantee the appropriate functions of the application. The task is to devise a TSCH schedule that synchronizes packets generated between various applications while satisfying each QOS standard. In order to deal with these requirements, not only minimizes slot frame sizes and delays, but also filling the required QOS parameters. In contrast to pr e-research, which focuses only on minimizing delay [15, 19] and maximizing reliability [11, 29, 30], this paper has multiple specified wireless sensor network. Treats the task of satisfying the QOS requirements.< S 3 , S 5 , S 6 >Another limit of the current TSCH scheduling algorithm is the premise of a fixed packet rate, and does not take into account the uniformity of packet rates, which are often characterized by rea l-world applications. Traditional research has often overlooked the impact of heterogeneous packet rates in the industrial sensor network, as explained in [9, 10, 11, 12, 13]. These studies generally focus on fixed packet rates, simplifying the subsequent transmission of each sensor node. These approaches may not be enough in scenarios, where packet rates change significantly, do not take into account the complexity caused by traffic changes in the actual industrial environment.

4. Methodology

Definition of data flow is essential for implementing and maintaining QOS policies. By associating QOS parameters such as latency and packet loss requirements to a specific data flow, you can ensure your application performance goal. This is very important in industrial environments where reliable and predictable communication is indispensable for operation efficiency and safety. In the following paragraphs, we will explain in detail how these flows are defined and how the application is assigned to these flows.

The wireless sensor network's Ochigi Topology is represented by graph G = (S, L), S represents a set of sensor nodes in S 1, s 2, ..., s, and n represents the total number of sensor nodes. , L represents a set of links. The sensor node is located in a square lattice, and the distance between the adjacent nodes shall be D I S T N BR. Assuming that all nodes have a uniform communication range, they are represented by R. This topology is composed of one sink node (S 1) and N T transmitted nodes (ST), and each node is associated with an application that has different QOS requirements. Each application is defined by a specific packet rate (PR) and is associated with two QOS metrics (target delay and allowable packet loss). In addition, a specific packet rate is assigned to each application. The target delay of the application I and the packet loss are TD I and T Li, respectively. Since the intensive TSCH scheduling approach is adopted, it is assumed that the sink node is recognized as the target QOS value after the first information exchange. Since the application is assigned to the transmission node, multiple flows will be created based on the number of routes to the sink node from each transmission node. Define the flow as a function F (S i): S I ∈ S T. The main purpose is to meet the specified QOS requirements for each flow. Figure 1 shows an example of a 6-node tree topology.

As shown in FIG. 1, the sink node is set to S 1, and the five transmission nodes are s t =. Definition of data flow is essential for implementing and maintaining QOS policy. By associating QOS parameters such as latency and packet loss requirements to a specific data flow, you can ensure your application performance goal. This is very important in industrial environments where reliable and predictable communication is indispensable for operation efficiency and safety. In the following paragraphs, we will explain in detail how these flows are defined and how the application is assigned to these flows.

Since the application is assigned to the transmission node, multiple flows will be created based on the number of routes to the sink node from each transmission node. Define the flow as a function F (S i): S I ∈ S T. The main purpose is to meet the specified QOS requirements for each flow. Figure 1 shows an example of a 6-node tree topology.

4.1. Phase 1: Population Initialization

As shown in FIG. 1, the sink node is set to S 1, and the five transmission nodes are s t =. Definition of data flow is essential for implementing and maintaining QOS policies. By associating QOS parameters such as latency and packet loss requirements to a specific data flow, you can ensure your application performance goal. This is very important in industrial environments where reliable and predictable communication is indispensable for operation efficiency and safety. In the following paragraphs, we will explain in detail how these flows are defined and how the application is assigned to these flows.

As shown in FIG. 1, the sink node is set to S 1, and the five transmission nodes are s t =.

In this scenario, two different applications, which are labeled with application 1 and 2, are introduced. The group of the sending node that processes the packet of application 1 is identified as ST 1, and the node that sends the application 2 packet is labeled ST 2. In the figure, five flows are shown: f (s 2), f (s 3), f (s 4), f (s 5), and f (s 6). Node

∈ S T1 sends an application 1 packet to Sync Node S 1.

∈ sends the packet of application 2 to S 1. The main purpose is to guarantee that each flow meets the specified QOS requirements related to each application.

As shown in Fig. 2, the TSCH schedule is composed of a matrix containing a channel offset representing individual cells and a time slot offset, and a timelot group forms a slot frame. Here, the slot frame consists of four time slots and has four channel offsets. The cell is indicated by the tu m e s long (T m m e s l o f s e-t, c h that n e lo f s e-t), and can be shared with multiple transmissions or for one transmission. One time slot will provide enough time for the transmitter to send the largest packet, respond to it and send a confirmation response. The network coordinator is responsible for the management and control of traffic flows, and calculates the optimized time slot and channel assignment based on specific purposes.

The coordinator node regularly broads the Enhanst Beacon (EB), including the current absolute slot number (ASN). ASN is the total number of time slots that have passed since the network deployment, as shown in Fig. 2.

In the QMDE development process, we set the following goals:

Considering the packet rate of the node, achieving the specified throughput requirements.

Follow the delay standard set in advance so that all nodes delayed within the defined limit.

Achieves the packet loss rate below or below the specified QOS threshold, improving the reliability of the entire network.

Figure 3 is a flowchart that shows the general process of QMDE in detail. This process was divided into two phases to clarify. Phase 1 focuses on the initialization of the mothe r-i n-law, and the phase 2 is specialized in the stage of the optimization process mutation, intersection, and selection stage. In the following, each phase is explained in detail.

Step 1: First, a set of random vectors (initial groups) of N pieces is generated. Each group is represented as P-P I, and is a vector of the size of Mx's. The p o p i consists of a value between V A R M IN and V A R MA X. Here, V Ar M IN is assumed to be 1, and V A R MA X is determined as a total period of transmitting calculated in the following formula:

V a r m a x = ∑ i = 2 n ⌈ l s f-pr (s) -D (s i) ⌉,

Here, PR (S I) represents the packet rate of the sensor node S i, and the D (S i) represents the depth of the sensor node S i in the tree structure. Furthermore, L S F is the slot frame length and is calculated by equation (2):

4.2. Phase 2: Mutation, Crossover, and Selection

L s f = l t s-n ts,

L s f = l T s-n ts, here, assuming that LT S is a time slot length and a standard value of 10 ms, NT S represents the number of time slots in the slot frame.

In this phase, we will explain how the early population for optimization is created and converted to a TSCH schedule through various steps. It also explains in detail how to calculate the module that identifies sensor nodes that can be added to the time slot without causing collisions or interference, and how to calculate the conformity of the schedule obtained as a result. The following paragraphs will explain the outline of the steps drawn in FIG.Step 1: First, a set of random vectors (initial groups) of N pieces is generated. Each group is represented as P-P I, and is a vector of the size of Mx's. The p o p i consists of a value between V A R M IN and V A R MA X. Here, V Ar M IN is assumed to be 1, and V A R MA X is determined as a total period of transmitting calculated in the following formula:

V a r m a x = ∑ i = 2 n ⌈ l s f-pr (s) -D (s i) ⌉,

L s f = l t s-n ts,

L s f = l T s-n ts, here, assuming that LT S is a time slot length and a standard value of 10 ms, NT S represents the number of time slots in the slot frame.

Finally, in this step, each P O PI is converted into an M × N matrix to match the dimension of the candidate TSCH schedule. The value of this matrix is used to map transmission, and eventually a candidate TSCH schedule is formed by the end of Step 5. Figure 3 is a flowchart that shows the general process of QMDE in detail. This process was divided into two phases to clarify. Phase 1 focuses on the initialization of the mothe r-i n-law, and the phase 2 is specialized in the stage of the optimization process mutation, intersection, and selection stage. In the following, each phase is explained in detail. In this phase, we will explain how the early population for optimization is created and converted to a TSCH schedule through various steps. In addition, we will explain in detail how to calculate the module that identifies sensor nodes that can be added to the time slot without causing collisions or interference, and how to calculate the conformity of the results obtained as a result. The following paragraphs will explain the outline of the steps drawn in FIG.

V a r m a x = ∑ i = 2 n ⌈ l s f-pr (s) -D (s i) ⌉,

L s f = l t s-n ts,

L s f = l T s-n ts, here, assuming that LT S is a time slot length and a standard value of 10 ms, NT S represents the number of time slots in the slot frame.

5. Performance Evaluation

Step 2: The TSCH schedule is built based on a pool containing a sensor node with a packet prepared for transmission. The pool is set individually for each time slot. The concept of the pool was introduced to save a qualified sensor node in each time slot that can be scheduled. A qualified node is a node that contains a packet in the queue, and is intended to be broadcast to a node or a sink node. By using the concept of this pool, the QMDE algorithm guarantees that the node in charge of the packet is not scheduled until at least one packet from the child node or the packet is generated. It is extremely easy to schedule a packet without a packet, listen unnecessary in the scheduled time, and monitor which sensor has a packet prepared to send to avoid deteriorating energy efficiency. It is important.

In the example shown in Fig. 1, assuming that all packets are generated by the initialization of the network, between the first time slot with the empty pool, S 2, S 3, S 4, S 5, and S 5, and S. 6 is eligible to be added to the first time slot pool. As a result, as shown in FIG. 4, these five nodes form a pool of a time slot 1.

If the pool of each time slot starts from an empty state, the sensor node is assigned to the TSCH schedule in the previous time slot, and there is no remaining packet in the queue, it will not be added to the pool in the next time slot. Conversely, the sensor node that has not yet been assigned moves to the pool with the next time slot until the queue is no longer the remaining packet. As shown in FIG. 5D, S 5 and S 6 are assigned to the time slot 1. As a result, as shown in FIG. 4, it is not included in the pool of the time slot 2. In addition, Node S 2 and S 3 with the expected two packets are added to the pool of the time slot 2. These two nodes will play the role of relaying packets from the child node for sink nodes, in addition to the transmission of their produced packets. Next, S 3 and S 4 are assigned (Figure 5D) in the time slot 2, and the pool is updated. This process can be continued until the schedule is no longer sent. Figure 4 shows an example of six pool statuses for six time slots.

Each time slot PO L S I Z E is supported by the total number of packets that can be sent in the time slot.

5.1. Simulation Setup

Step 3: In this step, the value of the column PO PI corresponding to a specific time slot in the TSCH schedule is normalized based on the specific time slot P-OL S I Z E. If the cell value of the matrix exceeds the pool size value during the normalization process, the value is replaced by the cell value module of the PO L S Iz E. Due to normalization, the value of each time slot in the P-O P i matrix fits in the same range as the relevant pool size.

Examples are shown to explain the normalization process better. In Fig. 5A, a 4x6 matrix is generated randomly in step 1. The value from 1 to 8 of this matrix indicates the total number of packets transmitted by network toologi drawn in FIG. 1. In FIG. 5A, the value of the time slot 1 must be normalized to 5 (POL S I Z E in the time slot 1) as shown in FIG. 4. Similarly, the value of the time slot 2 must also be normalized to 5. The values of these two time slots are not changed because they are already equal to the corresponding pool size or less. However, in the time slot 3, as shown in FIG. 4 (PO L S I Z E = 4), the expected number of transmission is 4, not 5. After normalization, the value of the time slot 3 in the queue drawn in FIG. The value of 4 or more is normalized. In other words, 5 is normalized to 1 (5m o d 4), and the remaining values are not changed. The normalization process is continued for time slots 4, 5, and 6, and the queue value is normalized based on the pool size of each corresponding time slot. The value shown in FIG. 5A is normalized as shown in FIG. 5B.

Step 4: After normalization, in Step 4, assign the sensor node to the TSCH schedule. This assignment is achieved by mapping the normalized matrix of each time slot in FIG. As shown in FIG. 5B, the value of the time slot 1 shown as 4, 5, 2, and 3 shows S 5, S 6, S 3, and, as shown in FIG. It is mapped to S 4 and supports the line number of the pool shown in Fig. 4 in the time slot 1. This mapping process is repeated to each time slot.

Step 5: In this step, each node assigned to a specific cell tries to include all matching pairs (MP) in the pool in that specific cell. A matching pair refers to a node in the pool, which does not cause a collision or interference that has already been scheduled. The addition of a matching pair can continue until the matching pair that can be added to the time slot TS is gone.

Updating the pool and its size value, regularization, mapping, adding matching pairs, and applying nodes are repeatedly executed to each time slot until the pool becomes empty.

Step 6: After building a schedule for each population, determine the cost of each schedule. Calculate the costs for the generated schedule from the viewpoint of packet loss. The QMDE algorithm generates an essential configuration file distributed to the node in the network. You can use the actual sensor network, but our case uses a network simulator. The simulator evaluates the provided schedule based on various service quality metrics. After execution of the simulator, the results indicate delay, throughput, and packet loss value, providing insights on QoS performance of the network by the generated schedule. The result of this step is a vector that includes delays and packet losses of each flow F (S i).

If any of these groups meets the QOS requirements of the application, the QMDE algorithm ends. If not, proceed to the phase 2 of the QMDE algorithm.

Phase 2 focused on optimization of population generation using an improved version of customized differential (CDE) algorithm. The CDE algorithm was initially introduced in [32]. However, QMDE has made further improvements, especially in the initialization and selection steps of the mother group. These adjustments are to generate a better schedule and select a group that can achieve QOS targets for applications defined in QMDE. Step 5: In this step, each node assigned to a specific cell tries to include all matching pairs (MP) in the pool in that specific cell. A matching pair refers to a node in the pool, which does not cause a collision or interference that has already been scheduled. The addition of a matching pair can continue until the matching pair that can be added to the time slot TS is gone.

Updating the pool and its size value, regularization, mapping, adding matching pairs, and applying nodes are repeatedly executed to each time slot until the pool becomes empty.

5.2. Simulation Results

5.2.1. Experiment 1: Achieving the QoS Values

If any of these groups meets the QOS requirements of the application, the QMDE algorithm ends. If not, proceed to the phase 2 of the QMDE algorithm.

Phase 2 focused on optimization of population generation using an improved version of customized differential (CDE) algorithm. The CDE algorithm was initially introduced in [32]. However, QMDE has made further improvements, especially in the initialization and selection steps of the mother group. These adjustments are to generate a better schedule and select a group that can achieve QOS targets for applications defined in QMDE. Step 5: In this step, each node assigned to a specific cell tries to include all matching pairs (MP) in the pool in that specific cell. A matching pair refers to a node in the pool, which does not cause a collision or interference that has already been scheduled. The addition of a matching pair can continue until the matching pair that can be added to the time slot TS is gone.

Updating the pool and its size value, regularization, mapping, adding matching pairs, and applying nodes are repeatedly executed to each time slot until the pool becomes empty.

If any of these groups meets the QOS requirements of the application, the QMDE algorithm ends. If not, proceed to the phase 2 of the QMDE algorithm.

5.2.2. Experiment 2: Slotframe Size

5.2.3. Experiment 3: Delay

CDE algorithms include the adaptation of the conventional differential evolution (de) algorithm [33] to effectively manage the inconsistent variables and variables that are unique to TSCH scheduling. This adaptation was extremely important due to the TSCH function that supports simultaneous transmission in one time slot or cell. The CDE generates a TSCH schedule by applying crossovers and mutations and mapping values to nodes without collision and node without interference.

QMDE not only assigns no n-collision and no n-interference to specific cells or time slots, but also uses pool concepts. This method makes it easier to generate schedules based on packets prepared for scheduling, and can avoid overscharming and reduced energy efficiency. Furthermore, the selection process is enhanced to support multipurpose optimization, guaranteeing that various QOS metrics are effectively satisfied.

5.2.4. Experiment 4: Reliability

In phase 1, n-PO P individual groups were generated.

5.2.5. Experiment 5: Time Complexity

As shown in FIG. 3, a mutant operation is performed when the calculated degree of compatibility of the multipurpose function of the generated maternal division of the generated population does not meet the request. , As follows, a candidate mother group called V I is generated.

V I = P O PR 1 + f × (P O PR 2-P O PR 3),

Here, R 1, R 2, and R 3 are no n-equivalent random numbers, are used to select up to three random groups from all individuals and generate candidate groups as V I from these groups. Function F is a mutant coefficient.

5.2.6. Experiment 6: Delay Comparison between QMDE and TASA

For each maternal group, QMDE generates a new maternal group V I. The mutation step generates a new solution candidate by adding a small correction to the existing solution.

Thereafter, by crossover operation, the algorithm mixes the ingredients from both the original group and the group V I, which is mutated, and generates a new group as N E W P O PI. Here, the components from the candidate group v i are replaced with a certain probability PC R (cross rate), and the compatible ingredients from PO PI are replaced with a certain probability of PO P i.

5.2.7. Experiment 7: Packet Delivery Ratio Comparison between QMDE and TASA

If it is not RA N D J ≤ P C R or j = I p o p i, it will be P O P i = V I.

6. Discussion

Here, RA N DJ is a random number of each component J.

After crossover, the generated population goes through pool setting, normalization, mapping, and MP pair addition (steps 2-5) to generate a candidate solution N e w P o p i . If the new candidate solution N e w P o p i proves to be better than the current solution P o p i , it replaces the existing solution. A better solution refers to a solution with smaller values of F D and F L as shown in equation (5). Otherwise, if the newly generated candidate is not better, the optimization continues until one of the following stopping criteria is met: (1) the number of iterations exceeds a defined maximum value M a x I t e r or (2) the generated schedule meets the primary objective set by the application.

The multi-objective optimization problem used in the QMDE algorithm is shown in equation (5). The scope of this optimization problem can be expanded to accommodate additional applications in the network with more QoS requirements. Since two QoS requirements are considered in this paper, there are two objectives (F D and F L) as shown in equation (5).

7. Conclusions and Future Work

F D = ∑ i ∈ S T 1 f ( S i ) . d - T D 1 size ( S T 1 ) + ∑ i ∈ S T 2 f ( S i ) . d - T D 2 size ( S T 2 ) , F L = ∑ i ∈ S T 1 f ( S i ) . l s - T L 1 size ( S T 1 ) + ∑ i ∈ S T 2 f ( S i ) . l s - T L 2 size ( S T 2 ) , O b j = M i n i m i z e

The first objective, denoted as F D , is to meet the delay requirements of the nodes running application 1 ( S T 1 ) and application 2 ( S T 2 ). The maximum target delays of these two applications are denoted as T D 1 and T D 2 , respectively. The variable f ( S i ) . d represents the end-to-end delay of node S i . F D calculates the average difference between the delay of the sending node in the candidate schedule and the delay required by each participating application.

Similarly, F L is defined in accordance with the same approach in accordance with the average difference between the packet loss of the node that runs application 1 and 2 in the candidate schedule and the target packet loss of the relevant application. It is formed to meet the packet loss requirements. Application 1 and 2 target packet loses are Tl 1 and T L 2, respectively. The packet loss of the sensor S I in the candidate schedule is represented by F (S I). L S. The main goal is to minimize the value of FD and F L, and the purpose is to reduce the difference between the delay of each fluff and the target value of the packet loss and the actual value.

In order to enhance the effectiveness of the optimization process and capture a wider range of solutions, the solution is the optimal candidate for the solution that is delayed or PDR improvement even if other metric deteriorates. Specifically, for example, if the PDR is improved and the delay deteriorates, or if the worsening does not exceed the preliminary stake, it is considered a candidate. These thresholds are set to 5 milliseconds for delay and 15 % for PDR.

Author Contributions

This section explains the outline of the simulation setup and gives a comprehensive evaluation of the QMDE approach. In particular, we focus on fluctuations in the number of nodes and traffic rates, and explore the effects of various factors on network performance. By adjusting the ratio of nodes assigned to a specific application, the network traffic rate is modulated and the overall impact on the overall performance is analyzed.

Funding

The implementation of QMDE was realized by co-simulation using Matlab and TSCH-SIM to get more accurate results. In addition, the TSCH-SIM simulator [35] has been enhanced by incorporating manual routing and scheduling functions into the TSCH-SIM simulator [34]. The TSCH schedule can be tested in various configurations by manually setting the TSCH schedule and static routing in the TSCH-SIM simulator. Details of manual routing and scheduling implementation in TSCH-SIM are described in [35]. Similarly, F L is defined in accordance with the same approach that evaluates the average difference between packet losses of nodes that run applications 1 and 2 in the candidate schedule and the target packet loss of related applications. It is formed to meet the packet loss requirements of some applications. Application 1 and 2 target packet loses are Tl 1 and T L 2, respectively. The packet loss of the sensor S I in the candidate schedule is represented by F (S I). L S. The main goal is to minimize the value of FD and F L, and the purpose is to reduce the difference between the delay of each fluff and the target value of the packet loss and the actual value.

Institutional Review Board Statement

Informed Consent Statement

Data Availability Statement

The implementation of QMDE was realized by co-simulation using Matlab and TSCH-SIM to get more accurate results. In addition, TSCH-SIM simulator [35] has been enhanced by incorporating manual routing and scheduling functions into the TSCH-SIM simulator [34]. The TSCH schedule can be tested in various configurations by manually setting the TSCH schedule and static routing in the TSCH-SIM simulator. Details of manual routing and scheduling implementation in TSCH-SIM are described in [35]. Similarly, F L is defined in accordance with the same approach in accordance with the average difference between the packet loss of the node that runs application 1 and 2 in the candidate schedule and the target packet loss of the relevant application. It is formed to meet the packet loss requirements. Application 1 and 2 target packet loses are Tl 1 and T L 2, respectively. The packet loss of the sensor S I in the candidate schedule is represented by F (S I). L S. The main goal is to minimize the value of FD and F L, and the purpose is to reduce the difference between the delay of each fluff and the target value of the packet loss and the actual value.

Conflicts of Interest

Abbreviations

The implementation of QMDE was realized by co-simulation using Matlab and TSCH-SIM to get more accurate results. In addition, the TSCH-SIM simulator [35] has been enhanced by incorporating manual routing and scheduling functions into the TSCH-SIM simulator [34]. The TSCH schedule can be tested in various configurations by manually setting the TSCH schedule and static routing in the TSCH-SIM simulator. Details of manual routing and scheduling implementation in TSCH-SIM are described in [35].	Matlab was used for the optimization process, and TSCH-SIM was adopted as a network simulation for determining QOS values. All processes to improve the TSCH schedule to obtain the optimal solution are performed offline. With this approach, the network does not stop during the optimization process, and the network will continue to work without interruption while determining the optimal schedule. Figure 6 shows a detailed sequence diagram of this c o-simulation. See [36] for more information about TSCH-SIM and Matlab's co-simulation.
Поначало уы сайы оазом васазой дой сде. еду соеди поеововог, в рат чозаиаяаная даная. з доздан п п оот юого соро са уа уа ка ка уогот о м е е норое соеопож п сорос паков, а з зощ я я ыом уом оадющ форац оорос паков посаной норац дов. сосав расан, го расасасан сди сди сов иов оое ообогого пого пого пду рам посан пакач паков и эосом нас ран рас памамамамамам т пож, каждое которых сой собый сабор тован Qos.	In order to evaluate the proposal algorithm, a co-simulation of MATLAB and TSCH-SIM [36] was conducted. The configuration of the simulation is distributed in a square gridtopology with a 16 to 100 nodes with a certain communication range of R = 30 m. The parameters and their range are defined in Table 1. The sink node was located in the center of the grid, and the minimum spanning tree (MST) algorithm was adopted to generate the sensor network critpology.
As noted in the previous section, the transmission node is associated with the application class. Each application includes the defined packet rate and the required QOS. The QOS requirements specified in each application match the guidelines outlined in the ISA SP100 standard [37, 38, 39]. This standard classifies industrial applications into six categories, which mainly focuses on class 2 (closed loop monitoring control) and class 4 (state monitoring).	In class 2 applications such as PL C-based surveillance systems, PLC (Programmable Logic Controller) sends commands to actuators and executes tasks. In discrete manufacturing, tasks are often completed on a sequence. Therefore, if the communication link between the supervisor (PLC) and the actuator is delayed, the production speed will be directly affected. In other words, the longer the waiting time for these communications, the slower the entire production process. However, applications related to class 4 are not so important in ratenshi. Class 4 typical applications consist of sensing devices and monitoring devices, such as even t-based maintenance applications.
Application 1 imitates a close d-loop control system (class 2), and is often critical in terms of delay. Application 2 imitates class 4 status monitoring applications. Application 1 requires a delay of less than 50 ms and a packet loss of less than 10-7. Application 2 functions as a state monitoring application, the delay requirement is less than 100 ms, and the target packet loss rate is 10-6 or less. In this type of application, the system is used to collect data and transfer to a server. For example, a node is placed to collect data such as temperature over a specific period. The collected data is useful for lon g-term temperature control strategies. The specifications of each application are clearly summarized in Table 2.	The packet rate was classified into the following two groups, and a specific packet rate group was assigned based on the applications executed by each node. As shown in Table 2, the sensor node that executes application 1 is assigned to the middle rate category (m), and the sensor node that executes the application 2 is assigned to a lo w-rate category (L).
DE	Table 3 shows five major scenarios used to evaluate the performance of QMDE algorithm. Each scenario is evaluated in the range of various sensors from 16 to 100. As a result, four subcategories occur in each scenario. Therefore, our approach is tested in a total of 20 unique scenarios.
EP	In the optimization part of the QMDE algorithm, we implemented the initial optimization parameters shown in Table 1. MaxIter represents the maximum number of iterations, and nPop represents the initial population number. The size of the decision variables of the differential evolution is equal to the maximum size of the TSCH schedule (m × n), where m is a fixed value of 4 channels, and n is the slotframe size, which is set to 500. It is important to emphasize that the selection of 500 time slots as the maximum slotframe size does not mandate the use of all time slots, but rather provides a range of random numbers from which to choose. For example, if the pool is empty by time slot 10, the remaining 490 time slots will not be utilized and will be removed.
Va r m i n is set to 1, and V a r m a x represents the total number of transmissions determined by equation (1). In Table 1, the crossover probability (PCR) determines the likelihood of using a crossover operation to recombine elements from the current population to generate a new population. A higher PCR favors exploration, while a lower PCR favors exploitation. We used 0. 8 because we want to increase the probability of generating new solutions. The scaling factor, defined by b e t a m i n and b e t a m a x , controls the adjustment of the difference vector during mutation. A smaller scaling factor strengthens local search, while a larger one broadens the search space.	Table 4 shows the parameters and corresponding values considered in the simulations.
LI	In this section, we present the experimental results of QMDE using Matlab and TSCH-SIM co-simulation. We evaluated several key metrics, including delay, reliability, slot frame size, and time complexity. Furthermore, we compared these results with those obtained with the TASA algorithm, focusing especially on delay and PDR (Packet Delivery Ratio). This comparison helps us understand how QMDE performs in different scenarios compared to TASA.
Network delay refers to the total time that the packet travels from the source node to the destination node. Here, the delay is evaluated by taking the time difference from the generation of the packet to the normal receipt of the root node. On the other hand, reliability shows the ability of a network that normally transmits data between the sender and the recipient, and is often quantified using en d-t o-end PDR.	Figure 7 shows the difference between delays in candidate solutions in scenario 5 using 64 nodes and PDR. If either delay or PDR difference is improved in the repetition of each optimization, the solution is considered a candidate if the other metric is slightly deteriorated. This explains the fluctuation seen in Fig. 7. If neither indicators show any improvements, the optimization process will continue until the termination standard is met.
In Fig. 7, the repetition 1 begins with a schedule showing a delay difference of almost 100 ms and a 9 % PDR. This schedule is replaced with a better one in repetition 2, the difference in PDR (4 %), and the difference in delay is almost the same. As mentioned above, if the delay is deteriorated, for example, if the delay deteriorates, or the vice versa, the solution will be recognized as a candidate if the deterioration is within the scope of a prior defined value. 。 These thresholds are set to 5 milliseconds for delay and 15 % for PDR.	In repetition 3, the difference in delay decreased by 25 ms, but the difference in PDR increased 14%. In repetition 3, the schedule discovered is eligible as a candidate solution due to the improvement of the delay difference, despite the fact that the increase in PDR gaps has fallen below. Eventually, as shown in Fig. 7, the QMDE algorithm is successfully identified by repetitive 8, and as a result, the QOS target defined for the two applications is obtained from the optimal solution. The difference in the achievement QOS value is zero.
MP	The same goes for all other scenarios, excluding scenarios involving more than 100 sensor nodes. In these scenarios, it was impossible to achieve the specified delay of the application. Considering that the duration of the time slot is 10 ms, the maximum allowable delay of the node that runs application 1 is equivalent to a 5-time slot, and the maximum delay in the node that runs application 2 is a 1 0-time slot. Equivalent to. As a result, it is impossible to find a transmission arrangement that satisfies these restrictions.
The size of the slot frame has a significant effect on the delay and is determined by the total number of time slots in the slot frame. According to Fig. 8, it is presumed that achieving QOS defined for the application will be affected by two major factors. As you can see in the scenario, the higher the ratio of nodes with high packet rates, the number of time slots required to achieve PDR or delay goals increases in an exponential manner. If you do not find the optimal schedule in the optimization process, increase the slot frame size until you meet the termination standard.	The delay of each flow is defined as a time difference from the transmission of the packet to the reception in the sync node. As mentioned earlier, the expected delays for the sensor that execute application 1 must not exceed 50 MS, and the sensor that runs application 2 must maintain the maximum delay to 100 MS. As shown in Fig. 9, the QMDE algorithm has succeeded in satisfying both applications for both applications in scenarios containing 16, 36, and 64 sensor nodes. However, if the network size is 100 nodes, the obtained delay value is slightly higher than the requirements of the considering scenario. Considering that the time slot is 10 ms, the maximum delay that is acceptable is equivalent to five time slots in the node that runs application 1, and 10 time on a node that runs application 2. Equivalent to a slot. As a result, it is quite impossible to find a transmission arrangement that satisfies these restrictions.
We have observed that the type of application run on a node close to the sync node plays an important role. Each application has a specified packet rate, and if a node close to the sink runs a higher packet rate application, it will lead to a higher delay. Such nodes have to send their own packets while enduring the burden of traffic from other nodes, making it difficult to manage traffic and achieve the specified delay. < SPAN> The size of the slot frame has a significant effect on the delay and is determined by the total number of time slots in the slot frame. According to Fig. 8, it is presumed that achieving QOS defined for the application will be affected by two major factors. As you can see in the scenario, the higher the ratio of nodes with high packet rates, the number of time slots required to achieve PDR or delay goals increases in an exponential manner. If you do not find the optimal schedule in the optimization process, increase the slot frame size until you meet the termination standard.	The delay of each flow is defined as a time difference from the transmission of the packet to the reception in the sync node. As mentioned earlier, the expected delays for the sensor that execute application 1 must not exceed 50 MS, and the sensor that runs application 2 must maintain the maximum delay to 100 MS. As shown in Fig. 9, the QMDE algorithm has succeeded in satisfying both applications for both applications in scenarios containing 16, 36, and 64 sensor nodes. However, if the network size is 100 nodes, the obtained delay value is slightly higher than the requirements of the considering scenario. Considering that the time slot is 10 ms, the maximum delay that is acceptable is equivalent to five time slots in the node that runs application 1, and 10 time on a node that runs application 2. Equivalent to a slot. As a result, it is quite impossible to find a transmission arrangement that satisfies these restrictions.
We have observed that the type of application run on a node close to the sync node plays an important role. Each application has a specified packet rate, and if a node close to the sink runs a higher packet rate application, it will lead to a higher delay. Such nodes have to send their own packets while enduring the burden of traffic from other nodes, making it difficult to manage traffic and achieve the specified delay. The size of the slot frame has a significant effect on the delay and is determined by the total number of time slots in the slot frame. According to Fig. 8, it is presumed that achieving QOS defined for the application will be affected by two major factors. As you can see in the scenario, the higher the ratio of a hig h-packet rate node, the number of time slots required to achieve the PDR or delay goals increases in an exponential manner. If you do not find the optimal schedule in the optimization process, increase the slot frame size until you meet the termination standard.	The delay of each flow is defined as a time difference from the transmission of the packet to the reception in the sync node. As mentioned earlier, the expected delays for the sensor that execute application 1 must not exceed 50 MS, and the sensor that runs application 2 must maintain the maximum delay to 100 MS. As shown in Fig. 9, the QMDE algorithm has succeeded in satisfying both applications for both applications in scenarios containing 16, 36, and 64 sensor nodes. However, if the network size is 100 nodes, the obtained delay value is slightly higher than the requirements of the considering scenario. Considering that the time slot is 10 ms, the maximum delay that is acceptable is equivalent to five time slots in the node that runs application 1, and 10 time on a node that runs application 2. Equivalent to a slot. As a result, it is quite impossible to find a transmission arrangement that satisfies these restrictions.
We have observed that the type of application run on a node close to the sync node plays an important role. Each application has a specified packet rate, and if a node close to the sink runs a higher packet rate application, it will lead to a higher delay. Such nodes have to send their own packets while enduring the burden of traffic from other nodes, making it difficult to manage traffic and achieve the specified delay.	The QMDE algorithm shows the effectiveness of identifying the optimal schedule that satisfies the specified PDR. In a specific scenario, algorithms may require additional time to identify the schedule, but as shown in Fig. 10, you can surely find the schedule. In the 16 and 3 6-node scenario, the optimal schedule achieves a 100 % complete PDR. However, as the number of nodes increases to 64 or more, this PDR value decreases slightly. This algorithm explores various solution candidates, but some take priority to minimizing delays, while others prioritize maximizing PDR. It is difficult to balance these conflicting purposes in a scenario. Nevertheless, the optimization process can continue until the optimal schedule is identified while balancing the PDR and the delay.
PR	As shown in Fig. 11, increasing the number of nodes and increasing the ratio of high packet rate to increase the number of repetitions and time to find the optimal schedule for the QMDE algorithm. In the 16-36 node scenario, the probability of packet loss due to queue overflow and high delay is minimal. This is because these scenarios have a small number of sending and smaller slot frame sizes. As a result, the generated schedules should be able to meet the QOS requirements, and there is no need to repeat them many times to find the optimal schedule. As the number of nodes increases, it is difficult to find the optimal schedule. Since there is no accurate information about the timing of the node generating packets, it takes time to assign an appropriate time slot without exceeding the expected delay or reducing the desired PDR. This task continues in a scenario with a high ratio of nodes sent at a high packet rate.
FIG. 11 shows that the time and complexity of the qmde optimization algorithm are exponential to the number of nodes. However, due to the nature of offline, there is no need to interrupt the network or repeat the algorithm repeatedly. When the optimal TSCH schedule is determined, it is distributed between nodes, so efficient operation is guaranteed without continuous adjustment.	It is important to consider the overhead associated with the proposed QMDE algorithm, but it should be clarified that this process is performed only once and needs to be recomputed only if the topology of the network changes. The computational complexity of the QMDE algorithm includes three main components: population initialization, mutation and crossover operations, and fitness evaluation. The population initialization requires setting up n P o p individuals, each of which is represented by a D-dimensional matrix. This step requires O ( n P o p - D ) operations. The mutation and crossover operations are applied to each individual of the population, each of which requires O ( n P o p - D ) operations per generation. The fitness evaluation in TSCH-SIM depends on the average packet rate A v g P R of the sensor node, the total number of nodes N, and the simulation time T s . Assuming that the fitness function evaluation has a complexity of O(simulator) per iteration, the computational complexity of these components together is O(nP o P - D - Simul a t o r - Max I t e r ).
Regarding latency, Figure 12 shows the average latency observed when running two applications (application 1 and application 2) simultaneously using the QMDE and TASA approaches. The average latency provides an overview of the overall latency performance across all defined flows in the network. The figure clearly shows that TASA exhibits consistently higher latency values compared to QMDE. Moreover, as the number of nodes increases, the average latency increases significantly for TASA, especially when the percentage of nodes running application 1 (which has a higher packet rate than application 2) becomes larger.	TASA assumes that all packets are generated during the network initialization step and prioritizes the schedule to nodes with a large number of packets in the queue, resulting in an unbalanced schedule when two applications with different QoS requirements are running. Since accurate packet generation times are not accessible in real networks, nodes may be scheduled before their packets are ready, resulting in increased delays as they wait until the next scheduled transmission slot. The QMDE algorithm optimizes the TSCH schedule and reduces delays by dynamically adjusting transmission allocations through pool updates, crossovers, and mutations. By optimizing the schedule for each application separately, QMDE identifies the optimal time slot for each transmission to efficiently meet the target delay of each flow.
As shown in Figure 13, QMDE maintains a high PDR, although it shows a slight degradation in scenarios with a large number of nodes. In contrast, TASA prioritizes scheduling to nodes with a large number of packets, which may result in packet loss at nodes with buffer overflows or long delays. TASA exhibits high PDR for small numbers of nodes (16 and 36 nodes), but this value decreases when increasing the number of nodes and network traffic. QMDE optimizes the scheduling process by iteratively adjusting the schedule to ensure that each application achieves the required PDR. This optimization involves strategically assigning sensor nodes to appropriate time slots to ensure the required number of nodes to meet the target PDR.	To evaluate the performance of the algorithm in various scenarios, extensive experiments and analysis were conducted using network simulators, but they have limitations when compared to real-world network conditions. For example, simulators like TSCH-SIM often simplify the distribution of configuration information, and schedules are typically specified in a single file that is distributed to all nodes. However, in real environments, this process is complex and is managed by IEEE 802. 15. 4e beacons. Such simplifications can lead to discrepancies between simulated and real-world performance.
Like all intensive algorithms, QMDE algorithm requires the coordinator nodes to have a complete topology information and the sensor node traffic rate. The network coordinator is responsible for managing and controlling traffic flows, calculating optimized time slots and channels, and distribution of schedules to the entire network. IEEE 802. 15. 4E Beacons usually guarantee that the topology of the intensive coordinator will continue to be updated. In addition, if a network change occurs, these updates will be transferred to the coordinator node via beacon, and the coordinator nodes will decide whether to continue the same schedule with minor changes or regenerate new optimal schedules. Masu.	From the time complex perspective, the algorithm was evaluated in various scenarios. However, scenarios containing more than 100 nodes offer issues in a typical dynamic environment to industrial networks. It takes a lot of time to regenerate the optimal solution and apply a new schedule to a scenario with more than 100 sensor nodes, and during this r e-optimization period, the network can be maintained, but this is possible. Temporary periods can lead to delays and losses of packets due to cuoverflows caused by changes in network topology and traffic patterns. In future research, it is possible to explore more efficient r e-optimization technology to deal with these issues.
In this study, we have introduced an approach in consideration of a new QOS that uses an extended multipurpose differential evolution optimization algorithm in order to design the optimal TSCH schedule for different species in the industrial environment. The purpose was to develop an algorithm that meets the QOS requirements of multiple applications that operate simultaneously in the network. Our approach has dealt with potential collisions and interference to achieve high PDR. In order to manage nodes with a wel l-prepared packet, we implemented a dynamic pool concept, minimized redundant scheduling, and reduced delays that could adversely affect PDR values. < SPAN> As with all concentrated algorithms, QMDE algorithm requires the coordinator nodes to have a complete topology information and the traffic rate of the sensor node. The network coordinator is responsible for managing and controlling traffic flows, calculating optimized time slots and channels, and distribution of schedules to the entire network. IEEE 802. 15. 4E Beacons usually guarantee that the topology of the intensive coordinator will continue to be updated. In addition, if a network change occurs, these updates will be transferred to the coordinator node via beacon, and the coordinator nodes will decide whether to continue the same schedule with minor changes or regenerate new optimal schedules. Masu.	From the time complex perspective, the algorithm was evaluated in various scenarios. However, scenarios containing more than 100 nodes offer issues in a typical dynamic environment to industrial networks. It takes a lot of time to regenerate the optimal solution and apply a new schedule to a scenario with more than 100 sensor nodes, and during this r e-optimization period, the network can be maintained, but this is possible. Temporary periods can lead to delays and losses of packets due to cuoverflows caused by changes in network topology and traffic patterns. In future research, it is possible to explore more efficient r e-optimization technology to deal with these issues.
In this study, we have introduced an approach in consideration of a new QOS that uses an extended multipurpose differential evolution optimization algorithm in order to design the optimal TSCH schedule for different species in the industrial environment. The purpose was to develop an algorithm that meets the QOS requirements of multiple applications that operate simultaneously in the network. Our approach has dealt with potential collisions and interference to achieve high PDR. In order to manage nodes with a wel l-prepared packet, we implemented a dynamic pool concept, minimized redundant scheduling, and reduced delays that could adversely affect PDR values. Like all intensive algorithms, QMDE algorithm requires the coordinator nodes to have a complete topology information and the sensor node traffic rate. The network coordinator is responsible for managing and controlling traffic flows, calculating optimized time slots and channels, and distribution of schedules to the entire network. IEEE 802. 15. 4E Beacons usually guarantee that the topology of the intensive coordinator will continue to be updated. In addition, if a network change occurs, these updates will be transferred to the coordinator node via beacon, and the coordinator nodes will decide whether to continue the same schedule with minor changes or regenerate new optimal schedules. Masu.	From the time complex perspective, the algorithm was evaluated in various scenarios. However, scenarios containing more than 100 nodes offer issues in a typical dynamic environment to industrial networks. It takes a lot of time to regenerate the optimal solution and apply a new schedule to a scenario with more than 100 sensor nodes, and during this r e-optimization period, the network can be maintained, but this is possible. Temporary periods can lead to delays and losses of packets due to cuoverflows caused by changes in network topology and traffic patterns. In future research, it is possible to explore more efficient r e-optimization technology to deal with these issues.

References

In this study, we have introduced an approach in consideration of a new QOS that uses an extended multipurpose differential evolution optimization algorithm in order to design the optimal TSCH schedule for different species in the industrial environment. The purpose was to develop an algorithm that meets the QOS requirements of multiple applications that operate simultaneously in the network. Our approach has dealt with potential collisions and interference to achieve high PDR. In order to manage nodes with a wel l-prepared packet, we implemented a dynamic pool concept, minimized redundant scheduling, and reduced delays that could adversely affect PDR values.
Verification by Matlab and TSCH-SIM co-simulation framework confirmed that the schedules generated by the QMDE algorithm effectively met the QoS requirements of diverse application flows. However, in networks with more than 100 nodes, it proved difficult to achieve the desired latency due to constraints in scheduling transmissions within a single time slot with a fixed number of channel offsets.
Comparing the latency and PDR of QMDE and TASA in various scenarios for selected applications according to ISA SP100 (closed-loop supervisory control and condition monitoring), the results revealed that QMDE consistently achieved lower latency and met the QoS objectives than TASA. QMDE's iterative optimization approach aims to identify the optimal time slots to allocate transmissions to meet the targeted latency and PDR requirements. Although TASA showed remarkable PDR and low latency in scenarios with fewer nodes (16 and 36), it struggled as the network traffic and number of nodes increased. This limitation stems from TASA's challenge to optimize time slot allocation and node assignment in the TSCH schedule to consistently meet delay and PDR requirements.
Future work will explore ways to improve the effectiveness of the QMDE algorithm in computing and updating the schedule in response to dynamic network changes, including topology changes, node additions, removals, and other events. This investigation aims to increase the adaptability and efficiency of the algorithm to maintain an optimal scheduling strategy amid evolving network conditions. In addition, we plan to investigate the impact of increasing the number of channel offsets in scenarios involving large networks, aiming to evaluate how effectively this strategy can reduce latency in large networks experiencing high traffic loads.
Supervision, R. L., Conception and methodology, A. V. and R. L., Software, Verification, Investigation, Formal analysis, A. V. and R. L., Writing (original draft), A. V., Review and editing, R. L. and A. V., Funding acquisition, R. L. All authors have read and agreed to the published version of this manuscript.
This research was supported by the Natural Sciences and Engineering Research Council of Canada Discovery, grant number RGPIN-2019-04454.
Not applicable.
Not applicable.
The data are included in the paper.
The authors have stated that there is no conflict of interest. The funding agency is not involved in the design, data collection, analysis, interpretation, manuscript writing, and announcement of results.
This manuscript uses the following abbreviations:
ASN
Absolute slot number
Amus
Adaptive mult i-hop scheduling
CDE
Customized difference evolution
Cmab
Combination mult i-arm bandt
Difference evolution
Expected packet
FDMA
Frequency split multiple access
Local repetition
LLR
Linear learning compensation
Mac
Media access control
Matching pair
MST
Minimum spanning tree
OSCAR
Optimization scheduling cellularocation algorithm
PCDe
Priorit y-based customized difference evolution
PDR
Packet reaching rate
Packet rate
QOS

Elim Poon - Journalist, Creative Writer

Last modified: 27.08.2024

The Frog Story slot by EGT comes with 5 reels and 20 paylines. The slot's theme is the tales of the Frog Princess. Hundreds of casino game studios offer various games: from traditional titles like video slots, poker, roulette, and bingo to live dealer games and virtual. Online Casino Featured Slot Frog Story by Amusnet - Gaming Provider. Play Free spins are currently available on most of our portfolio of slot games.

Play

accepted

Frog Story Slot Machine Free Online Casino Game by EGT

Frog Story

Once Upon A Time…

Venturing Further Into The Forest

Magic Is In The Air

Final Spell

A Magical Game For All

Frog Story Slot FAQs

Free to Play EGT Slot Machine Games

EGT Slot Machine Reviews (No Free Games)

You Might Also Like

Casino Bonuses

Online Gambling

Mobile Slots

Live Dealer Casinos

Our Favourite Casinos

Claim Free Spins, Free Chips and Much More!

Quality of Service-Aware Multi-Objective Enhanced Differential Evolution Optimization for Time Slotted Channel Hopping Scheduling in Heterogeneous Internet of Things Sensor Networks

1. Introduction

2. Related Works

3. Problem Definition

4. Methodology

4.1. Phase 1: Population Initialization

4.2. Phase 2: Mutation, Crossover, and Selection

5. Performance Evaluation

5.1. Simulation Setup

5.2. Simulation Results

5.2.1. Experiment 1: Achieving the QoS Values

5.2.2. Experiment 2: Slotframe Size

5.2.3. Experiment 3: Delay

5.2.4. Experiment 4: Reliability

5.2.5. Experiment 5: Time Complexity

5.2.6. Experiment 6: Delay Comparison between QMDE and TASA

5.2.7. Experiment 7: Packet Delivery Ratio Comparison between QMDE and TASA

6. Discussion

7. Conclusions and Future Work

Author Contributions

Funding

Institutional Review Board Statement

Informed Consent Statement

Data Availability Statement

Conflicts of Interest

Abbreviations

References