A decentralized algorithm to combine topology control with network coding
Introduction
Network coding comprises two distinct sub-problems: coding and sub-graph selection. The coding sub-problem determines each intermediate node should do which operation. The sub-graph selection argues on the routes between the sources and destination nodes, which should be optimal in cost. Since network coding, specifically sub-graph selection, needs to know about the neighbors (this information is used to make the routes), topology control is in conflict with the network coding opportunity [20], [25]. In fact, topology control uses some relay nodes to assist transmitting the packets while in network coding the relay nodes may mix the packets with an operator and then send the coded packets [1], [34].
In the following, we illustrate an example of exchanging data in a WSN where topology control and network coding are used, simultaneously. See Fig. 1, where the nodes and want to exchange their own packets, and , respectively. The first choice is direct transmission where the nodes and exchange their packets directly with no relay node (see Fig. 1-a). However, the nodes and can send their packets to the relay node that mixes the packets and then sends. The nodes and can retrieve each others’ packets by XOR-ing with their own packets (see Fig. 1-b). Note that without network coding, transmissions are required because the node should send and independently. This transmission can be accomplished by the relay node , too. Here, the node exhausts more energy to communicate with the farther node , while the node uses less transmission range. Additionally, both nodes and may be employed as relays in which either or mix the packets. The best solution for this problem is affected by the various factors, such as generated traffic of nodes, the distance between them, and their residual energy.
Recently, some studies, see e.g. [18], [22], [23], have been introduced which combined network coding with topology control. These studies provided some mathematical theories and showed some advantages, such as increasing the lifetime, and decreasing energy consumption. To the best knowledge of authors, available approaches use either a heuristics method or a centralized solver to solve the problem. However, the heuristics methods obtain a sub-optimal solution (generally is not close to the optimal solution) and the centralized approaches are subject to some well-known disadvantages. By considering the advantages of combining network coding with topology control, a decentralized approach is required in which the nodes obtain the network topology and network-coding-based routes, independently.
In this paper, a convex optimization problem is proposed that considers multiple sources and combines network coding with topology control. Then, Lagrange dual, sub-gradient, and the decomposition methods are utilized to present a decentralized algorithm. The proposed algorithm repeatedly solves the simpler sub-problems until it achieves the global object that is the optimal lifetime. We assess the performance of the proposed approach through simulations and compare the obtained results with the available optimal solutions. The simulation outcomes show that considering multiple sources leads to more lifetime compared to considering the sources, separately. We also study the effect of network parameters on the convergence rate of the proposed algorithm. The main contributions of this paper are summarized as follows:
- •
It considers multiple sources and provides a convex optimization problem for optimal topology control in network-coding-based-multicast WSNs. The mathematical formulation can help us to assess the essence of this problem.
- •
For the first time, a decentralized and optimal algorithm is presented that combines network coding and topology control. The proposed algorithm creates an automated and dynamic structure for WSNs, which continuously changes the routes and transmission range of the nodes.
- •
We prove the efficiency of the proposed algorithm by comparing it with the optimal solution and state-of-art approaches.
The rest of this paper is organized as follows: Section 2 presents an overview of related literature. System model and problem definition are described in Section 3. Section 4 describes the proposed decentralized algorithm. The assessment of approaches is done through simulations and the results are reported in Section 5. Finally, conclusions and directions for future research are presented in Section 6.
Section snippets
Related work
Both network coding and topology control techniques provide various advantages, such as extending the lifetime, saving energy, reducing interference, decreasing Media Access Control (MAC) and routing complexity [34]. This section reviews the studies that consider these techniques simultaneously. Note that various topology control mechanisms (e.g. power assignment, transmission range adjustment, and communication resource allocation) are considered in these studies.
The authors of [26], [31]
Problem statement and formulation
This section presents the system model, then develops an optimization problem for combining network coding with topology control. The proposed optimization model considers multiple sources in a WSN and finally it is converted into a convex non-linear programming.
Suppose that a set of wireless nodes, , is randomly deployed in the monitoring environment and consists of the sensors, , and sinks, (i.e. ). Formally, the hyper-graph defines the set of nodes, , and their connections by
Proposed decentralized algorithm
There are polynomial-time and practically fast algorithms that can solve the convex optimization (8)–(13). However, the computational characteristics of WSNs do not allow us to use the centralized algorithms. More clearly, it is almost impossible or impractical to gather all information to solve the proposed optimization problem, especially when the number of deployed sensors is large and network dynamically changes. The main advantage of the proposed optimization problem is its decomposability
Performance evaluation
This paper considers controlling the topology of network-coding-based WSNs with multiple sources and proposed a convex programming and a decentralized algorithm which are called OpMuToNe and DeMuToNe, respectively. In this section, simulation setup and results are presented that show the performance of these approaches.
To evaluate the performance of studied approaches, we consider that the nodes are randomly deployed in a square region. The number of sensors and sinks are equal to
Conclusion
This paper addresses network lifetime optimization in a WSN that employs both network coding and topology control, jointly. The proposed convex optimization problem considered multiple sources in a network in which can provide higher efficiency. However, obtaining its solution faces some essential difficulties especially for large networks. By utilizing Lagrange dual, sub-gradient and decomposition methods, a decentralized and iterative algorithm was developed which obtains optimal lifetime for
CRediT authorship contribution statement
Moammad Khalily-Dermany: Conception and design of study, Acquisition of data, Analysis and/or interpretation of data, Writing - original draft, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Mohammad Khalily-Dermany is currently an assistant professor in the Department of Computer Engineering at Islamic Azad University, Khomein branch. He received his B.Sc in 2002 from Islamic Azad University, M.Sc. in 2006 form, Amirkabir University of Technology (Tehran Polytechnic) and Ph.D. from Islamic Azad University, Science and Research branch, Tehran, Iran, all degrees are in computer engineering. His research interests include network coding, analysis of wired and wireless sensors,
References (34)
- et al.
Energy-efficient dynamic resource allocation with opportunistic network coding in OFDMA relay networks
Comput. Netw.
(2012) - et al.
Network coding-based reliable multicast in wireless networks
Comput. Netw.
(2010) - et al.
A survey on network coding: From traditional wireless networks to emerging cognitive radio networks
J. Netw. Comput. Appl.
(2014) - et al.
A multi-channel cooperative MIMO MAC protocol for clustered wireless sensor networks
J. Parallel Distrib. Comput.
(2014) - et al.
Network coding-based energy-efficient multicast routing algorithm for multi-hop wireless networks
J. Syst. Softw.
(2015) - et al.
Itinerary planning for mobile sinks in network-coding-based wireless sensor networks
Comput. Commun.
(2017) - et al.
Combining topology control and network coding to optimize lifetime in wireless-sensor networks
Comput. Netw.
(2019) - et al.
A convex optimization model for topology control in network-coding-based-wireless-sensor networks
Ad Hoc Netw.
(2017) - et al.
Cross-layer topology design for network coding based wireless multicasting
Comput. Netw.
(2015) - et al.
A survey of inter-flow network coding in wireless mesh networks with unicast traffic
Comput. Netw.
(2015)
Random network coding
Energy aware resource allocation in multi-hop multimedia routing via the smart edge device
IEEE Access
Asynchronous Algorithms for Large-Scale Optimization: Analysis and Implementation
A distributed fault-tolerant topology control algorithm for heterogeneous wireless sensor networks
IEEE Trans. Parallel Distrib. Syst.
Constrained Optimization and Lagrange Multiplier Methods
Convex Optimization Algorithms
Cited by (6)
Transmission power assignment in network-coding-based-multicast-wireless-sensor networks
2021, Computer NetworksNetwork-Coding-Enabled and QoS-Aware Message Delivery for Wireless Sensor Networks
2023, Wireless Personal CommunicationsMulti-criteria itinerary planning for the mobile sink in heterogeneous wireless sensor networks
2023, Journal of Ambient Intelligence and Humanized ComputingA new optimal protocol for Green IoT communication
2023, Intelligent Green Communication Network for Internet of Things
Mohammad Khalily-Dermany is currently an assistant professor in the Department of Computer Engineering at Islamic Azad University, Khomein branch. He received his B.Sc in 2002 from Islamic Azad University, M.Sc. in 2006 form, Amirkabir University of Technology (Tehran Polytechnic) and Ph.D. from Islamic Azad University, Science and Research branch, Tehran, Iran, all degrees are in computer engineering. His research interests include network coding, analysis of wired and wireless sensors, resource allocation, optimization in constrained networks, and protocol design.