Congestion Control Based On Routing Optimization In WSNs

With the rapid development of wireless sensor networks (WSNs) technologies, ne-twork congestion control based on routing optimization has attracted more and more attention. Wireless sensor networks, with low power consumption, small size, multi-fu-nction and many other advantages in intelligent transportation, environmental monitori-ng and other fields, have been fully applied. However, due to that the wireless sensor network has a high error rate, being susceptible to interference, limited node energy, and so on. At the same time, the multimedia network development brings more traffic flow, making the congestion problem in wireless sensor networks increasingly promin-ent, and the network quality of service has been seriously affected. The router is resp-onsible for sending the information collected by a number of nodes to the sink node, which is the core technology of wireless sensor networks, and the importance is self-e vident. Wireless sensor network in routing optimization, including queue management and quality of service, has become a hot issue of network congestion control research.Based on the routing optimization problem, this article discusses wireless sensor network with the congestion control and network quality of service. For the network node’s energy consumption, delay, bandwidth and other indicators, according to a diff-erential-elite ant colony algorithm, this paper gets the system entropy, and applied the-m to optimize the multicast routing. The transfer function based on the classical cont-rol theory closed-loop system, aimed to seek algorithms of active queue management, to maintain queue stable in routing buffer. The main research works are as follows:(1) This paper first proposes a difference-elite ant colony algorithm. Before the r-outing protocol is established, getting the ACO"s optimized parameters by differential algorithm, obtaining the differential-ant colony system. while the ant’s pheromone-upd-ating strategy is optimized, the paper solved and proved pheromone consistency limit, introducing the "elite preservation strategy" from the genetic algorithms; at the same t-ime, for the wireless sensor network multicast routing optimization in different busine-ss streams with different QoS’s indicators of problem, a system utility function based on entropy is proposed; confirmatory proof of the algorithm was conducted by multip-le sets of simulation experiments. The network topology has been proposed based on a minimum of distance means algorithm.(2) In router’s optimization part with AQM, for the sensitive and dependence sh-ortcoming of RED algorithm with parameters selection, as well as relying artificial ex-perience, the paper acquired RED parameters combination based on classical control t-heory. Based on the study of classical control theory as well as the dynamic low fre-quency window and dynamic queue model, this paper analyzed Taylor expansion linea-rization to obtain a system differential equations containing pade delay, and thus getti-ng the congested network’s Mason flow diagram. On this basis, analyzing the open-lo op transfer function system on amplitude-phase frequency characteristics to handle the closed-loop stability, and determined the magnitude of the open-loop transfer function with the phase stability conditions, and applied them to RED algorithm’s parameter s-election. At the same time, in order to deal with the load disturbance on queue length, and to improve basic RED and get a fast response to load changes, proportional co-ntroller was designed. Meanwhile, dependent on the choice of parameters, the paper made a confirmatory experiment by obtaining the stability condition proposed. For act-ive queue management algorithm, the experimental simulation part was completed with network simulation software NS2, and conducted the simulation of traditional queue management algorithms of the drop tail algorithm (DropTail). The NS2’s comparative simulation with control law, demonstrated superiority in selecting RED parameters to maintain queue stability, meanwhile, comparing to the classic RED, using a proportion-al controller can effectively respond to load changes, transmission delay jitter, and ma-ke quick changes in disturbance response, reducing the impact caused by the queue 1-ength’s shock.(3) For the sensitive characteristics of wireless sensor networks, this paper put fo-rward a new controlling algorithm based on proportional-integral (PI) control+inner-1 oop negative feedback, and applied them to congested network with delay item of the object. The new method, based on adjustment of parameters in inner-loop, overcame the affects coming from disturbance of controlled parameters. Meanwhile, the introduc-tion of proportional integral controller eliminated the steady-state error. The comparing simulation experiments with the Simulink software showed the robustness with syste-m’s unit step disturbance, and gave a rational range with the inner-loop controlling p-arameters.