A wireless weak-connected network routing algorithm inspired by Physarum polycephalum

https://doi.org/10.1016/j.compeleceng.2021.107083Get rights and content

Abstract

Wireless weak-connected network (WWN) is a special self-organizing network, which can self-organize in an unstructured style and communicate with each other without infrastructure support. In this network, the link connection is weak and the topology is dynamic, which limits the performance of routing tasks. In order to address this issue, a novel Physarum-inspired routing algorithm (P-iRA) is proposed based on the intelligence and adaptability of Physarum polycephalum. Firstly, we construct a link capacity model for WWN via mapping the Physarum polycephalum network. Then, we design a next-hop selection strategy to obtain an optimal forwarding node in candidate set. Finally, we present a stochastic route strategy to achieve data transmission by virtue of finding the best routing path. Simulation results demonstrate that P-iRA improves approximately 47%, 34%, 56%, 20%, and 8% compared with Direct Delivery, Epidemic, First Contact, MaxProp, and Spray And Wait, respectively.

Introduction

With the rapid development of multi-hop communication between mobile terminals [1], [2], wireless networks of weak connection and dynamic topology draw increasing attentions, which have been widely used in special scenarios of harsh communication environment, such as remote area network, sudden disaster environment, and vehicle self-organizing network. However, most existing routing methods are difficult to tackle communication problems under weak connection and unstable topology. To this end, wireless weak-connected network (WWN) is proposed, which relies on the movement of forwarding nodes to form a transient connected sub-region to achieve best-effort communication and data routing. In this network, the duration time of communication link is quite short. Thus, we adopt the store-carry-forward mechanism, in which the selection of the optimal next hop node is the key issue in the routing research of WWN.

Currently, there are already some routing methods for complex network environments with weak connections and dynamic topology changes. They exhibit different network performance in terms of delivery ratio, network overhead, and transmission delay. To maximize delivery ratio, Shen et al. [3] proposed Epidemic routing algorithm based on the Lagrangian and dual problem models by using optimal buffer scheme. Tuan et al. [4] investigated a multi-copy routing strategy and a buffer management policy that optimize the delivery rate, forwarding a message to a neighboring node that has both a stronger social tie with the destination and a similar queue length. Literature [3], [4] enhances the packet delivery ratio in stochastic routing, but incur a heavy communication overhead. To make up these deficiencies of high overhead in the above works, Spyropoulos et al. [5] proposed Direct Delivery algorithm based on forwarding strategy, in which source node only delivers the message to the destination node. Jain et al. [6] presented the First Contact algorithm, where the node forwards message to the first neighbor it encounters, reducing effectively the network overhead. These algorithms reduce the resource overhead, but suffer from the maximum transmission delay. Based on the limitations of the above works, Burgess et al. [7] proposed MaxProp algorithm, where message is prioritized by using the link’s historical data and several complementary mechanisms. These methods shorten transmission delay via avoiding flooding messages, but it is difficult to set the message validity period. Additionally, although they improve only some aspect of performance, they lack of comprehensive performance consideration. Therefore, its application range is limited.

In recent years, artificial intelligence methods [8], [9] have become research hotspots, such as deep learning methods based on neural networks [10], [11], [12], [13], [14], [15] and bio-inspired intelligence methods [16]. A slime mold, Physarum polycephalum has become an emerging research hot spot due to its intelligent and self-adaptive behaviors of foraging process. When food sources are randomly distributed in the experimental area, Physarum can change its shape at will during the nutritional period and build a protoplasmic pipeline network connecting all food sources. In addition, there is a positive feedback mechanism for pipes and connectivity, which can adjust own pipes to change in network and automatically correct the wrong routing paths in real time. Especially, this pipeline network has been proved to be comparable with the real-world infrastructure network in terms of efficiency, fault tolerance, and cost [17]. As a consequence, we learn from this idea to solve the routing problem in WWN. Physarum polycephalum constantly finds the path is an idea of best-effort transmission, which can effectively increase the data delivery ratio; Physarum polycephalum adaptively optimizes the path selection, which can greatly reduce the network load; Physarum polycephalum can quickly adapt to the environment and automatically correct the wrong path, which can effectively reduce the path-finding delay.

In this paper, we propose a novel Physarum-inspired routing algorithm (P-iRA) for WWN based on intelligence and adaptability of Physarum foraging process, in order to solve delivery ratio, network overhead, and transmission delay. First, Physarum constantly seeking for path is used to enhance the data delivery ratio. Second, Physarum self-optimizing the path selection is used to reduce the network overhead. Finally, the rapid adaptation of Physarum to the environment is used to reduce transmission delay.

The main contributions of this paper are as follows: (1) Under the special communication background with weak connection and dynamic topology, WWN is proposed for the first time. (2) Due to Physarum foraging process is very similar to the communication process between nodes, it helps to construct a bio-inspired WWN network model. The physical quantity in the foraging model is migrated to the WWN through the dimensionless analysis method to solve the next hop selection problem of the routing. (3) A novel Physarum-inspired routing algorithm is developed by taking into account both the mobility stability of nodes and the quality of link communication, including the next-hop selection and stochastic route strategies. (4) Experiments are carried out and results confirm that three indicators of delivery ratio, transmission delay, and overhead ratio are good.

The rest of this paper is organized as follows. The related work is presented in Section 2. Section 3 builds a WWN network model based on Physarum foraging behavior. The adaptive WWN routing method is detailed in Section 4. Section 5 illustrates the experiments and results. This paper is concluded by Section 6.

Section snippets

Related work

Currently, there already exist many routing researches for weak connection and dynamic topology networks. Ayub et al. [18] proposed a lock routing protocol, which reduces message drop and curbs resource consumption. Sidera et al. [19] presented the extended minimum estimated expected delay (EMEED) protocol, which performs better in jointly optimizing the packet delivery rate. Zhang et al. [20] designed an opportunistic routing protocol to achieve a tradeoff between delivery efficiency and

Mathematical model of Physarum polycephalum

Physarum polycephalum has the characteristics of migration and fusion, and shows a high degree of intelligence and adaptability [17]. Physarum polycephalum foraging network shows in Fig. 1. At t=0, a small plasmodium of Physarum polycephalum (yellow dot) is placed in the experimental arena within the white border, and supplement with additional food sources (white dots) in the region. Over time, the plasmodium grows out from the initial food source with a contiguous margin and progressively

Adaptive WWN routing method

Based on the model equation in Section 3, we present an adaptive WWN routing algorithm composed of two strategies. The next-hop selection strategy is to identify a suitable next-hop within the communication range of the forwarding node. The stochastic route strategy is to realize data transmission under weak connection.

Experimental setup

In this paper, we simulate the P-iRA algorithm by using the ONE [24] simulator, and compare the P-iRA algorithm with the Direct Delivery [5], Epidemic [3], First Contact [6], MaxProp [7], and Spray And Wait [25] algorithms by changing the simulation parameters. The detailed simulation parameter settings are given in Table 1.

Performance comparison

We set three aspects including the number of nodes, node buffer, and node moving speed and analyze their impact on the delivery ratio, network overhead ratio, and delivery

Conclusion

In this paper, using the inspiration from Physarum foraging for patchily distributed food sources, we propose P-iRA to solve the issue of data best-effort delivery under dynamic topology change and weak connection for WWN. The P-iRA adopts the mode of delivery while routing to ensure that data packets reach the destination end via next hop node. In the WWN network model, we obtain the function expression of link capacity and node adaptive selection through mathematical analysis. Then, we design

CRediT authorship contribution statement

Xiang Hua: Supervision, Investigation, Funding acquisition. Hongjuan Yao: Data curation, Writing - original draft, Conceptualization, Methodology. Zhao Wang: Data curation, Validation. Baohua Li: Formal analysis. Hai Wang: Software.

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.

Acknowledgment

This work is supported by Key Research and Development Program of Shaanxi province, China in 2020 [grant number 2020GY-073].

Xiang Hua received her M.S. and Ph.D. degrees from Wuhan University and Northwestern Polytechnical University respectively, in 2004 and 2011. Her current research interests include short distance data communication and wireless network detection control.

References (25)

  • SideraA. et al.

    Wireless mobile dtn routing with the extended minimum estimated expected delay protocol

    Ad Hoc Netw

    (2016)
  • LuH. et al.

    User-oriented virtual mobile network resource management for vehicle communications

    IEEE Trans Intell Transp Syst

    (2020)
  • LuH. et al.

    Drrs-bc: Decentralized routing registration system based on blockchain

    IEEE/CAA J Autom Sin

    (2020)
  • ShenJ. et al.

    Buffer scheme optimization of epidemic routing in delay tolerant networks

    J Commun Netw

    (2014)
  • TuanL. et al.

    A joint relay selection and buffer management scheme for delivery rate optimization in dtns

  • SpyropoulosT. et al.

    Single-copy routing in intermittently connected mobile networks

  • JainS. et al.

    Routing in a delay tolerant network

  • BurgessJ. et al.

    Maxprop: Routing for vehicle-based disruption-tolerant networks

  • LuH. et al.

    Brain intelligence: Go beyond artificial intelligence

    Mob Netw Appl

    (2018)
  • LuH. et al.

    Deep fuzzy hashing network for efficient image retrieval

    IEEE Trans Fuzzy Syst

    (2021)
  • ChenY. et al.

    Research on image inpainting algorithm of improved gan based on two-discriminations networks

    Appl Intell

    (2020)
  • ChenY. et al.

    Research on image inpainting algorithm of improved total variation minimization method

    J Ambient Intell Humaniz Comput

    (2021)
  • Cited by (1)

    Xiang Hua received her M.S. and Ph.D. degrees from Wuhan University and Northwestern Polytechnical University respectively, in 2004 and 2011. Her current research interests include short distance data communication and wireless network detection control.

    Hongjuan Yao received her B.S. degree from Xi’an Technological University in 2018. She is studying toward the M.S. degree in communication and information systems. Her current research interests include wireless sensor network topology control and artificial intelligence algorithm simulation.

    Zhao Wang received her B.S. degree from Communication University of China in 2018. She is working toward the M.S. degree in armament science and technology at Xi’an Technological University. Her current research interests include networks cooperative control and intelligent optimization algorithm.

    Baohua Li received his B.S. degree from Xi’an Technological University in 2019. He is working toward the M.S. degree in communication and information systems. His current research interests include wireless sensor network, path planning, and graph theory algorithm.

    Hai Wang received the B.S. degree in electrical engineering and automation from Xi’an Technological University in 2018. He is currently a assistant engineer with North Special Energy Group Xi’an QingHua Company. His research interests include wireless network and algorithm simulation.

    This paper is for regular issues of CAEE. Reviews processed and approved for publication by the co-Editor-in-Chief Huimin Lu.

    View full text