Design and optimization of node architecture for application specific multi-hop wireless networks.

摘要

In this dissertation, we focus on the design and optimization of the node architecture for target specific applications under the 802.11-based multi-hop wireless network that has been recently spotlighted from the public domain. We consider two target applications: compression and packet aggregation algorithms on the Wireless Mesh Network (WMN), and a multi-modal tracking system underlying the multi-hop wireless network.;For the tracking system, we develop a multi-modal sensor-based tracking model where acoustic sensors mainly track the target objects and visual sensors compensate the tracking errors. We initially discover a network synchronization problem caused by the different location and traffic characteristics of multi-modal sensors, and non-synchronized arrival of the captured sensor data at a processing Server. We show the improved tracking accuracy from visual compensation in ideal case is severely degraded when the synchronization problem is involved in real situations. For the possible solution for the problem, we differentiate the service level of sensor traffic based on Weight Round Robin (WRR) scheduling at the Routers. The weighting factor allocated to each queue is calculated by a proposed Delay-based Weight Allocation (DWA) algorithm. In addition, to numerically predict the number of success of the visual compensation, we propose a Statistical Estimation Algorithm (SEA) which is based on traffic measurement, random generation of transmission delay for sensor data, and statistical estimation. Based on the SEA, we propose an on-line version of the SEA called Statistical Estimation and Adaptation Algorithm (SEA2), in which the acoustic sensor's object sampling interval is automatically adapted to achieve a certain level of tracking accuracy.;To conduct the design and optimization tasks of the compression and packet aggregation, we propose a profile-based network and hardware co-simulation method to characterize the global WMN performance as well as the real-timing nodal behaviors. The WMN is equipped by a dedicated hardware platform possibly configured as a network processor. For the compression, RObust Header Compression (ROHC) is adopted. The co-simulation method integrates the network level simulator, NS-2 and hardware level simulator, SystemC. In this approach, we first insert the modules of ROHC and packet aggregation algorithms into the network simulator hierarchy, and measure the packet arrival times. Then, the corresponding hardware architecture is designed by SystemC for profiling the hardware delay appeared in encoding and de-coding packets. Finally, the traced hardware delays are applied into the network level simulator to extract real-timing WMN behaviors changed by the hardware operation in each mesh router. Additionally, to accurately predict the capacity of a hardware design, we propose a numerical analysis method by using the open Jackson queueing network. The modeled queue systems are one-to-one mapped into the constructed hardware components to characterize the concurrent operations and interactional relationship between encoding and decoding paths in ROHC.