认知无线视觉传感网络机会传输的分布式跨层优化

为了充分利用空闲授权无线电频段和增强视觉信息的端到端传输质量，研究了认知无线视觉传感网络机会传输的跨层设计问题。在分析信道随机性和网络模型的基础上，把跨层设计问题表达为一个视觉信息峰值信噪比和网络平均传输时延的权衡优化问题。通过对该问题进行对偶分解和基于随机次梯度的求解方法，提出了一个分布式跨层传输优化算法。该算法不需要预先知道可用授权频段的静态概率分布，而通过节点在每个时隙中进行独立计算和局部信息交换使得上层视觉感知信息的压缩速率与底层链路机会传输自适应匹配，达到权衡优化问题的最优解，因此可以作为认知无线视觉传感网络的实用传输协议。仿真结果表明，该分布式算法能够快速收敛，并能获得与集中式最优化算法相似的性能。

Distributed Cross-Layer Optimization for Opportunistic Transmission in Cognitive Wireless Visual Sensor Networks

To fully utilize the vacant licensed radio bands to enhance the transmission quality of visual information, a cross-layer design for the opportunistic transmission of cognitive wireless visual sensor network was studied. A trade-off optimiza-tion between peak signal-to-noise ratio and the average transmission delay of visual information was formulated based on the analysis of stochastic channel model and network model. By applying dual decomposition and stochastic sub-gradient method to the dual problem, a distributed cross-layer optimization algorithm was proposed. Without the knowledge of the stationary probability distribution of the licensed bands, the algorithm can achieve an adaptive matching between the com-pression rate of the perceived visual information in the upper layer and the opportunistic transmission of the links in the lower layer through independent computation and local information exchanging in relevant nodes in each slot, thus obtaining the optimal solution to the trade-off optimization problem. The proposed algorithm can be used as a practical transmission protocol in cognitive wireless visual sensor networks. Simulation results show that the distributed algorithm is able to con-verge quickly and achieve the performance similar to that of the optimal centralized algorithm.