Research on topology reconstruction and key technologies of UAV trunking communication network *

Wang can, Gao Chenzi (the 28th Research Institute of China Electronics Technology Corporation, Nanjing 210007)
Label (space separated): paper reading

• UAV cluster operation is the trend of future war. In the process of attacking the enemy, due to the mobility of nodes in the UAV cluster communication network and the enemy's fire attack, the links between nodes are often disconnected, resulting in the damage of communication network connectivity. Aiming at the connectivity repair of UAV trunking communication network, this paper studies the classical methods of topology reconstruction and the advantages and disadvantages of each method, and further discusses the key technologies and future development trend of UAV trunking communication network topology reconstruction.
• Keywords UAV cluster; Topology reconstruction; Network connectivity; key technology

1. Research reasons

In the process of UAV cluster operation, it is bound to suffer from the fire attack of the enemy, resulting in the damage of some UAV nodes. Once the damaged node is the key node connecting the two communication network partitions, it will affect the connectivity of the whole UAV cluster communication network. How to repair the damaged area of UAV cluster communication network connectivity in real time and dynamically is a key problem to be solved in UAV cluster operation.

2. Research content

Aiming at the connectivity repair of UAV cluster communication network, the classical methods of * * * topology reconstruction * * * and the advantages and disadvantages of each method are studied, and the key technologies and future development trend of UAV cluster communication network topology reconstruction are further discussed.

3. Topology reconstruction of UAV trunking communication network (development status)

3.1 development status of UAV cluster

    2000 Defense Advanced Planning and research agency( DAR⁃PA)The research plan of UAV cluster combat technology was released［4］. In recent years, the United States has carried out a lot of research on UAV cluster and achieved some preliminary results. For example, DARPA "Elf project" announced in 2015［5］. The project plan aims to put a large number of miniaturized and recyclable UAVs outside the enemy's defense range, realize the cooperation among multiple UAVs by using wireless ad hoc network technology, suppress the enemy by attacking the enemy's basic communication facilities and security facilities, and even attack the enemy's database system by using computer viruses. In the field of UAV cluster control technology, the intelligent unmanned cluster project team led by researcher Zhao Yanjie of China Electronics Technology Corporation has made remarkable achievements in the world. On June 10, 2017, the team successfully completed the first cluster flight experiment of 119 fixed wing UAVs in the world, breaking the previous record of 103 fixed wing UAVs held by the US military.

3.2 development status of network topology reconstruction concept

    The concept of topology reconfiguration originated from distribution network in order to improve the fault tolerance of the whole distribution network［6］. Similarly, when the failure of nodes in the network affects the communication links in the network, it is also necessary to restore the performance of the network through various reconstruction technologies to meet the normal communication needs of the network. move ad hoc As a distributed and self-organizing wireless network, its nodes are mobile, the topology changes frequently, and the links in the network are easy to be disconnected. UAV trunking communication network is a typical mobile communication network ad hoc Network, each UAV is distributed and completing its own tasks.
    The nodes in the UAV trunking communication network are often damaged and invalid by the deliberate attack of the enemy. However, the failure of a large area of nodes or the failure of key nodes in the network will not only affect the network coverage of the combat area, but also divide the network into multiple disconnected partitions, affect the data exchange between nodes in the network, and even lead to network paralysis, which seriously affects the cooperative combat capability of the whole UAV cluster. Therefore, fault tolerance and partition connectivity restoration in wartime environment are the key issues of UAV cluster technology［7］. 
    The so-called topology reconfiguration refers to changing the topology of the network by controlling the mobile nodes or links in the network, so as to make the network have higher performance and better provide services for users. Topology reconfiguration is an effective means to restore network connectivity, that is, when network failure leads to network zoning, connectivity is damaged and performance is significantly reduced, run topology reconfiguration algorithm to restore network connectivity by supplementing new network nodes or redundant nodes in mobile network. According to the implementation strategy of topology reconfiguration, it can be divided into preventive reconfiguration and restorative reconfiguration. The basic idea of preventive reconfiguration is to know the network nature and topology in advance and take certain measures to prepare for possible network failures in the future. The basic idea of restorative reconfiguration is to take corresponding measures to restore the network connectivity when the network is divided due to failure. Compared with preventive reconfiguration, its timeliness is lower. Low effectiveness.
    The existing research on topology reconstruction technology mostly focuses on the field of wireless sensor networks, and there is less research on the topology reconstruction technology of UAV cluster communication network. Nodes in sensor networks are usually static nodes without mobility, which are powered by batteries in most cases. Node failure is usually caused by energy depletion or hardware failure. The nodes in the UAV cluster communication network move frequently, and the energy of a single operation is sufficient. Most of the reasons for its failure are due to enemy attacks, and the UAV network has higher requirements for the real-time recovery of connectivity. It is necessary to complete the topology reconstruction of the whole network in a very short time, restore the connectivity of the network, and ensure the survivability and cooperative combat ability of the UAV cluster. It is found that the application scenarios and requirements of UAV cluster communication network topology reconstruction and wireless sensor network topology reconstruction are different. The traditional wireless sensor network topology reconstruction method is difficult to be directly applied to UAV cluster communication network. Therefore, the research on the topology reconstruction method of UAV cluster communication network in battlefield environment is of great significance to improve the survivability and cooperative combat ability of UAV cluster.

4. Introduction to typical algorithms of topology reconstruction

4.1 topology reconstruction method based on power control

    The premise that the two nodes of the wireless link can communicate is the ratio of the power of the electromagnetic wave signal transmitted by the transmitting node to the interference and noise of the receiving node, that is, the signal-to-interference-noise ratio( Signal-Inter⁃face-Noise-Rate，SINR)It should be greater than the threshold of the receiver of the receiving node. By controlling the transmission power of the node, the number of neighbors within one hop of the node can be adjusted, and then the topology complexity of the whole network can be controlled. Power control is an important means to realize network topology reconfiguration.
    Although the power control technology can achieve the purpose of topology reconstruction to a certain extent, simply increasing the transmission power to re link the damaged link will increase the number of neighbors in the one hop range of nodes and aggravate the competition for wireless channel resources among nodes. Especially in the UAV cluster communication network with high node density, when the network service load is large, the conflict between nodes will increase sharply, which will seriously affect the network performance. Therefore, how to balance network connectivity and communication efficiency is the focus of topology reconstruction method based on power control.

4.2 topology reconstruction method based on node mobility

1)Network node role determination: the node determines its role according to the impact of its state change on the network topology, including ordinary nodes, redundant nodes and key nodes. Generally, the nodes that can be used to repair key nodes without affecting network connectivity are called redundant nodes, the nodes connecting different partitions are called key nodes, and other nodes are called ordinary nodes. 
2)Network fault diagnosis: according to the enemy's attack model and fault diagnosis algorithm, diagnose the current fault node type and its neighborhood nodes, and locate the fault source. Only when the key nodes in the network are damaged, the topology reconstruction process will be initiated to repair the network connectivity. 
3)Network topology repair: the repair node is usually served by redundant nodes. The best repair node is selected according to the adaptability function in the neighborhood of key nodes. The preset motion model of the repair node moves to the position of the key node. If the movement of the repair node causes a new network topology segmentation, cascade movement is carried out until the network is connected again.
4)Network topology optimization: the purpose of network topology reconstruction is to restore the connectivity of the network on the one hand and ensure the coverage of the network on the other hand. After the repair node completes the connectivity repair, it also needs to jointly optimize the network connectivity and regional coverage by taking the coverage of the node repair area in its neighborhood as the goal.

4.3 topology reconstruction method based on Clustering

    Such methods usually consider hierarchical wireless networks, in which all nodes move in an accessible open area. Nodes with large transmission range are assigned as cluster heads( cluster heads，CH)，And be responsible for collecting and forwarding messages. Other nodes are cluster members( cluster members，CM)，They are CH Randomly deploy around. CM Calculate their relationship with nearby CH Euclidean distance between and join the nearest cluster.

5. Analysis on Key Technologies of communication network topology reconstruction

    The network topology reconstruction method based on node mobility is the most suitable for UAV cluster communication network. However, at present, most of these topology reconstruction methods are proposed for wireless sensor networks, which is difficult to be directly applied to UAV cluster networks with large number of nodes, high density and frequent topology changes. At the same time, such methods still have many core technologies that need further breakthroughs, mainly including the following contents:

5.1 determination technology of network key nodes

    The large-scale UAV cluster communication network has a large number of nodes, and the global traversal methods such as depth search and breadth search are time-consuming and inefficient; The distributed search method based on one hop neighbor determines that the key node is the global key node, which has low probability accuracy. Balancing the accuracy of information acquisition and network load overhead is the key problem that key node algorithms need to pay attention to. How to design an efficient and accurate key node determination method according to the characteristics of UAV cluster is one of the key technologies to be studied in network topology reconstruction.

5.2 network fault diagnosis technology

    Network topology reconfiguration can be divided into two stages: reconfiguration trigger and reconfiguration implementation. In the trigger phase, it can be completed through topology fault diagnosis. In the implementation phase, it can be completed through topology reconstruction. Fault diagnosis is an important means to detect network anomalies and determine damaged nodes. The higher the efficiency of fault diagnosis, the smaller the overhead of reconfiguration trigger. Under the guarantee of efficient and reasonable reconfiguration implementation scheme, we can realize high real-time network topology reconfiguration and minimize the decline of network performance caused by the damage of network connectivity. At present, the research at home and abroad mainly focuses on wired networks, and there are many mature methods, such as diagnosis based on neural networks, fault diagnosis methods based on Bayesian theory and so on.

5.3 network connectivity recovery technology

    Network connectivity restoration methods are mainly divided into two categories: one is for mobile wireless networks, which completes the self restoration of network connectivity by replacing the failed nodes with the movement of redundant nodes in the network, and the other is to restore the network connectivity by placing additional relay nodes. UAV cluster combat system is a system with high real-time requirements. The method of placing relay nodes to restore network connectivity is difficult to be guaranteed. Therefore, the UAV trunking communication network is more suitable to replace the failed nodes through the movement of redundant nodes in the network to complete the self recovery of network connectivity. However, at present, the research of such methods mostly focuses on the failure of a single node. In the process of UAV cluster operation, it often encounters the deliberate attack of the enemy, resulting in the failure of a large area of nodes at the same time, which is difficult to deal with by the existing methods. To sum up, network connectivity repair technology is also the key technology to be studied in the topology reconstruction of UAV trunking communication network.