Ship Ad-hoc Network (SANET)

Professor : Dr.Aref Safari
November - December 2019
Faculty of Engineering
Islamic Azad university,Branch Rasht

1. Ship Ad-hoc Network (SANET) Student : Benyamin Moadab An undergradute Student of Computer Engineering (IT), At Islamic Azad University, Rasht Branch IN THE NAME OF GOD Professor : Dr.Aref Safari PhD in Artificial Intelligence November - December 2019 Faculty of Engineering

2. Next Page Table of Contents Introduction (Wireless sensor network - Multi-hop ) Wireless ad-hoc network (Diagram & Tabels) Ship ad-hoc network (SANET) Challenge Evolution Algorithm Optimization issues (Diagram & Tabels ) Compare methods Evaluation of algorithms Improve Suggestions Conclusion Reference

3. Types of Communication Networks What is Wireless Sensor Network? A Wireless Sensor Network (WSN) is a collection of compact size low power computational nodes capable of detecting local environmental conditions, and forward such information to a central base station (sink) for appropriate processing. “ 1 ) Infrastructure networks Ad-Hoc Network 2 ) 1

4. Ad-Hoc Network Applications MANETs - VANETs - SANETs - FANETs - SPANs - Air Force UAV 1- Lack of infrastructure 2- Use wireless link 3- Multiple jumps 4- Autonomy of nodes in relocation Three security methods in wireless networks 1- WPA / WPA2 / WPA3 / WEP 2- SSID 3- MAC Routing protocols 1- Table Driven Protocols 2-On Demand Protocols 3-Hybrid Protocols (or Hierarchical) 4-Multi Way Routing Routing Algorithms 1- Flooding Method 2-The Rumor spread method 3-Spin Method 4- Guided Release Method 5-Hierarchical Method Wireless ad hoc network Feature Of Ad-Hoc Network Ad-Hoc Network Problems 1- Routing 2- Optimal energy consumption 3- Quality of Service ( QoS ) 4- Multicast service delivery 5- Security Difference ? 2

5. 3

6. Type Of Routing Protocol Algorithm used Advantages and Disadvantages Tabel Driven DSDV Bellman-Ford ( Dijkstra's ) Requires time parameters (Update information and number) WRP Path-Finding The counting problem was resolved to infinity CSGR Bellman-Ford ( Dijkstra's ) Computational overhead Star - Can't find the optimal path and doesn't have the update required On- Demand SSR By power and signal Weak signal loss DSR - Have cache memory TORA Distributed Routing Algorithm ( eg : Ant Colony and Digestra , etc.) Requires synchronous clock AOPV Bellman-Ford ( Dijkstra's ) - RDMAR Spacing algorithm ( eg : distance-vector, LS, etc.) No need for positioning technology like GPS, Reduces network traffic 4

7. Method Of Routing Algorithm used The method how to worked ?! Hierarchical Method K-Means LBG ACO-MRP Single-Link LEACH Protocol PEGASIS HEED Protocol TEEN Protocol 5 This algorithm always produces k clusters with at least one node in them. This algorithm only optimizes the center of the clusters but does not optimize their boundaries. The K-Means algorithm depends on the initial selection of the clusters and this causes the clustering results to differ from the replication algorithm to solve this problem. In this algorithm a multi-routing protocol based on dynamic clustering and ant colony algorithm is proposed. This method is one of the oldest and simplest clustering methods and is one of the hierarchical and proprietary clustering methods. This method of clustering is also called the Nearest Neighbor technique. In this protocol the time is divided into parts called round. Each round is divided into two phases. The first phase is the cluster formation phase and the second phase is related to normal network performance . This method is essentially the advanced state of the LEACH protocol. In this protocol, instead of forming different clusters, a chain of interconnection is established between all the sensors in the network. The HEED clustering method is a distributed clustering method that considers both the energy and the cost of communication. This protocol is designed for situations where sudden changes in the parameter being measured are to be reported quickly

8. Vanet Ad-hoc Network There are two types of nodes in Vanet networks : 1) OBU (On-Board Unit) processing and communication equipment 2 ) Fixed Road-Side Unit ( RSU ) equipment As a result, these architectures form two types of communication: 1) Vehicle to Vehicle ( V2V ) 2) Vehicle to Infrastructure / Vehicle to Infrastructure ( I2V / V2I ) 6

9. Ship Ad-hoc Network Packet data networks at sea offer the potential for increased safety, connectivity and meteorological data acquisition. Existing solutions including satellite communication are expensive and prohibitive to most small vessels . ( Service Radio Regularization (RR ) defined marine band VHF radio to operate on internationally agreed frequencies in the band from 156MHz to 163MHz. ) 7

10. Challenge Important Challenges As we mentioned earlier, there are various problems and challenges in ad-hoc networks, the most important of which : Read More Security Routing 8

11. Evolution Algorithm In clustering, the main focus is on cluster head selection, but by adding the evolutionary methods described below, the clustering algorithm in the network selects the optimal set and optimal path. particle swarm optimization Differential evolution Genetic Algorithm 01 02 03 9

12. Function optimization 1. Single-Point and multi-Point optimization 2. Dynamic and static optimization Continuous or discrete optimization Optimization issues 3. 4. 10

13. Difficult To Optimize Optimization Diagram 11

14. Optimization Diagram Singel -point Multi-point Derivative-based Derivative - free (direct search) Steepest descent conjugate gradient Quasi-Newton Random Walk Hooke-Jeeves (Pattern search algorithm) Multi-Start and Clustering techniques Nelder-Mead Differential evolution * multi-point optimizers that operate on many points in parallel * multi-start algorithms that visit many points in sequence. 12

15. 01 02 03 04 Particle Swarm Optimization (PSO) ( Evolution Algorithm) 05 V i = w * V i (t-1) + c1 * rand1 * (P i.best – X i (t-1)) + c2 * rand2 * (P g.best -X i (t-1)) X i = X i (t-1) + V i (t) P i.best = (p i1, p i2, p i3, …. , p id ) P g.best = (p g1, p g2, p g3, …. , p gd ) X i = (x i1, x i2, x i3, …. , x id ) V i = (v i1, v i2, v i3, …. , v id ) 13

16. 01 02 03 04 Genetic Algorithm(GA) (Evolution Algorithm) 05 Population initialization Fitness Function Selection Crossover Mutation 14

17. 01 02 03 04 Differential Evolution (DE) (Evolution Algorithm) Population initialization Selection Crossover Mutation 15

18. Mutation / Crossover / Selection (Differential Evolution) 16

19. GA DE PSO 1)To solve small and big issues 2) Based on probability laws 3) Escape from the local optimal 4) Easy implementation 5) Lack evolutionary operator learn mode learn mode learn mode Compare methods 1)To solve just big issues 2) No mathematical proof 3) Easy implementation 4) High cost of information retention 5) Low speed calculation 1) Strong mathematical proof 2) High speed calculation 3) Easy implementation 4) Having memory 5) Discard less than GA 17

20. Assessment PSO GA DE 18

21. Improve Suggestions Fuzzy Particle Swarm Optimization : (in order to cluster suppliers in fuzzy environments) Integrating Fuzzy K-Means, Particle Swarm Optimization: (These algorithms help the FKM to escape from local Optima and improve the quality of solutions obtained by FKM) Other evolutionary hybrid methods .... 19

22. Conclusion According to surveys, satellite communications were expensive in communications between ships, For this reason they used networks called "case networks" to communicate between them over long distances. There were some methods already, but it was decided to use evolutionary algorithms in the field. 4 Suggested methods As you can see, they had good results and helped us optimize this 3 2 The main focus of the presentation was on routing algorithms 1 20

23. Reference Al- Zaidi , Rabab , et al. "Next generation marine data networks in an iot environment .“ 2017 Second International Conference on Fog and Mobile Edge Computing (FMEC) . IEEE, 2017 . Taher , Hadeel Mohammed. "Sea Ad hoc Network (SANET) Challenges and Development [J ]." International Journal of Applied Engineering Research 13.8 (2018): 6257-6261 . Kong, Peng-Yong, Ming- Tuo Zhou, and Jaya Shankar Pathmasuntharam. " A routing approach for inter-ship communications in wireless multi-hop networks ." 2008 8th International Conference on ITS Telecommunications . IEEE, 2008 . Verma , Rajesh, et al. "A hybrid wireless Ad-hoc network model for critical services ." 2010 Sixth International conference on Wireless Communication and Sensor Networks . IEEE, 2010. Course file "Differential Evolution Algorithm" by Dr. Rahil Hosseini 21