Cross-Layer Active Predictive Congestion Control Protocol for Wireless Sensor Networks

Abstract

In wireless sensor networks (WSNs), there are numerous factors that may cause network congestion problems, such as the many-to-one communication modes, mutual interference of wireless links, dynamic changes of network topology and the memory-restrained characteristics of nodes. All these factors result in a network being more vulnerable to congestion. In this paper, a cross-layer active predictive congestion control scheme (CL-APCC) for improving the performance of networks is proposed. Queuing theory is applied in the CL-APCC to analyze data flows of a single-node according to its memory status, combined with the analysis of the average occupied memory size of local networks. It also analyzes the current data change trends of local networks to forecast and actively adjust the sending rate of the node in the next period. In order to ensure the fairness and timeliness of the network, the IEEE 802.11 protocol is revised based on waiting time, the number of the node's neighbors and the original priority of data packets, which dynamically adjusts the sending priority of the node. The performance of CL-APCC, which is evaluated by extensive simulation experiments. is more efficient in solving the congestion in WSNs. Furthermore, it is clear that the proposed scheme has an outstanding advantage in terms of improving the fairness and lifetime of networks.

Resource Allocation Optimization for Heterogeneous Network

“…For the transmission of the data, optimization approach based on power interference controlling [1], cross layer coding [2] and proposed joint cross layer coding is carried out. The packet forwarding of the three methods are then recorded.…”

“…It is observed that the computation time for interference based controlling mechanism is very low than the other two method. However in the case of memory control mechanism [2], the time taken is higher due to the computation of memory buffering and optimization of allocation. However for the proposed approach, the processing time is still higher about 1.5 units, due to the optimization of multiple factors simultaneously.…”

“…Towards the development of congestion controlling at each node, a cross layer active predictive congestion control (CL-APCC) is proposed in [2]. This approach applies a network flow analysis of the memory status for each node in the network.…”

“…The proposed approach is as outlined below. For the allocated power of P i, derived from the interference model from [1], at node n i , considering the traffic blockage at, a node n i as B i , the observed probability of congestion level is defined as [2]…”

“…This approach presented a maximum power allocation problem into two simpler linear programs for symmetric and asymmetric links. In [2] with the similar objective to improve performance, a congestion avoidance approach is proposed. However no approach was observed controlling these two factors, to achieve higher performance.…”