Research On Resource Management Problems About The Wireless Self-organizing Network

With the development of Internet of Things and Ubiquitous communication technology, the single-hop communication mode which relies on the centralized control of the access point could not meet the demand of "connect everything". With the characteristics of distributed operation, networking flexibility and scalability, wireless self-organizing network has become a bridge which connects things and things, thing and people will be widely used in various fields.At present, the wireless self-organizing network derives a variety of new network forms. With the emergence of self-organizing edge computing network, self-organization participation sensing network and wireless sensor networks, the collaboration between these nodes in the network is gradually extended from communication to computing,storage resources and information resources. Network management is also extended from the allocation of wireless link resources to the management of computing resources, storage resources and information resources.With this trend, resource management is particularly important.In this paper, we comprehensively and systematically analyze the resource management problem in the wireless self-organizing networks and focus on the problems of network resource management, node resource sharing and collaborative management, computing resource management, information resources management and software-based wireless network management. In the following, we describe these resource management problems in details.1. Link resource management in traditional wireless self-organizing networks.Link resource management mainly refers to rationally integrate the wireless link resources in the network and design high-performance routing mechanism. As a key technology, routing protocol has been highly concerned by the researchers since the birth of the wireless self-organizing network. However, due to the characteristics of dynamic network topology and inter-node interference, there is a big challenge in routing design for wireless self-organizing network. The high stability routing mechanism for practical application still attract the attention of the researchers. In this part we study the stability routing mechanism for traditional wireless self-organizing networks, aiming at designing a high-stability routing mechanism based on node state prediction.The main work of this part is as follows: Firstly, the condition of the existence for the communication path and the condition of the packets can be transmitted on the path are analyzed. Then, the of link duration prediction algorithm is proposed based on the current movement of nodes,the probability factor of node congestion is proposed based on the length of node's buffer queue and the change rate of the length. Next, based on the information of link duration and the probability factor of node congestion, an on-demand routing mechanism for path reliability and stability is designed. Finally, we verify the performance of this routing mechanism through NS2 simulator.2. Link resource sharing and collaborative management of the nodeIn the wireless self-organizing network, the computing, storage and energy resources of the terminal are limited. These constraints cause the selfishness of the nodes in the network, that is, in order to save their own resources some nodes do not assist the other nodes to complete the communication task. The current network protocols are based on the collaboration of these nodes, the selfish behavior is one of the major bottlenecks which affect network performance. In order to solve the selfish problem in wireless self-organizing networks, we combine game theory and virtual currency mechanism to establish the cooperation enhancement model and design the cooperation enhancement mechanism.The main work of this part is as follows: First, the profits of the source node and the forwarding node in the packet transmission process are analyzed. The forwarding node consumes its own resources to forward the packet for the source node and obtain the virtual currency paid by the source node, the source node obtains the forwarding service and pays the corresponding virtual currency. Through the analysis, a cooperation enhancement model is established based on game theory. In this model, the network cooperation problem is transformed into a game equilibrium problem about the benefit of source node and forwarding nodes in the data forwarding path. In the meantime, in order to avoid the congestion and maximizing network lifetime, the model makes some certain constraint about the energy and storage space for the forwarding nodes. Then, we get the cooperation enhancement mechanism through solving the model. At last, the cooperation enhancement mechanism is analyzed.3. Computational resource management in self-organizing mobile edge computing networkWith the popularity of intelligent mobile terminals, the network communication mode has undergone tremendous changes, which mainly presents as the mobile access and mobile edge data exchange. Under this trend, the resources of cloud computing are distributing in the edge of the network, which results in mobile edge computing. In this part, we study the problem of computing resource management in self-organizing mobile edge computing and design the task scheduling mechanism.The main work of this part is as follows: Firstly, the task scheduling problem in self-organizing mobile edge computing system is analyzed.Secondly, the task scheduling problem is formulated as a distributed multi-device task scheduling game. Then, by constructing a potential function, the task scheduling game is proved to be a potential game,which possesses a property of finite improvement and always owns a Nash equilibrium. Next, the overhead-optimizing multi-device task scheduling algorithm is designed and the computational complexity is analyzed. Finally, simulations are conducted to evaluate the effectiveness of the proposed strategy. The results show that the proposed task scheduling strategy can effectively minimize the overhead of the mobile device and successfully complete the tasks.4. Information resource management in self-organized participatory sensing networkWith the development of information technology, modern smart devices are equipped with a lot of powerful embedded sensors, such as accelerometer, digital compass, gyroscope, global positioning system,digital compass etc. Using these embedded sensors, the public users act as participants to sense information from the surrounding environment and share the information through existing communication infrastructure,this forms a new research area which named participatory sensing. In order to facilitate the operation of the participation sensing system, we design an information resource sharing mechanism based on free competition theory for the self-organizing participatory sensing system.The main work of this part is as follows: firstly, we introduce the participatory sensing system and analyze the main problems about it. In the participatory sensing system, these participant share information and get the reward, we treat it as a free competition market and model the information sharing process as the Stackelberg game, the task publisher and the participants are both players in this game. Then, through analysis,it is proved that this game possesses a Nash equilibrium. Next, based on the above model, a platform-aided information sharing strategy is designed. The algorithm mainly involves three algorithms, which are the initialization algorithm, the collaboration enhancement algorithm and the interaction algorithm, then, the practicability and computational complexity of these algorithms are discussed. At last, the simulation platform is built to evaluate the performance of the proposed information sharing mechanism.5. Software definition wireless sensor networkWireless sensor network is a distributed self-organizing network,through the deployment of a large number of sensor nodes, we can obtain information in a certain area, which extends the antenna of the Internet to perceive the physical world and is an important source of network's data resource. With the large-scale application of the sensor network, the traditional sensor network exposes some problems: Different application scenarios have different requirements for sensor network. Sensor nodes need to be customized and it could not be used in other scenarios. The existing distributed control and routing algorithms bring a lot of calculation and information interaction, which increase the consumption speed of the energy and shorten the life time of sensor nodes and network.It is very difficult to implement congestion control and security policy on the portable, low-power sensor terminals.In order to solve the problems existing in the traditional sensor network, we introduced the concept of software defined network,adopting the idea of centralized control to realize the separation of numerical plane and control plane. Which can increase the flexibility to manage the wireless sensor network, effectively solve these high energy consumption and poor flexibility problems.The main work of this part is as follows: firstly, the composition,characteristics and characteristics of the traditional wireless sensor network are analyzed and the existing problems of the sensor network are pointed out. Then, by combining architecture of software defined network and the architecture of traditional wireless sensor network, we proposed a new architecture of wireless sensor network—Software Defined Wireless Sensor Network (SDWSN). Next, we put forward the design of network flow for the software defined wireless sensor network, which includes topology discovery, flow table creation, flow table distribution and topology maintenance. At last, we build a Linux-based hardware platform to test the performance of the proposed software defined wireless sensor network.