大规模数据并发环境下视频监控系统的研究与实现

城市轨道交通综合监控系统中的视频监控子系统与多个业务子系统共处同一个网络环境中，具有大规模数据并发特点。首先，介绍视频监控子系统所采取的策略，然后介绍基于订阅分发模型的视频数据规划，最后，重点研究视频分发过程中的负载均衡技术。在多台服务器工作的条件下，根据服务器的性能和负载等因素，提出一种基于动态测量的负载均衡技术，实验证明，该负载均衡技术可以有效地降低时延，提高系统的吞吐量，并均衡各个服务器的负载，取得良好的效果。

Research and Implementation of Video Surveillance System in Large Scale Data Concurrent Environment

The CCTV sub-system in the integrated supervisory and control system of urban rail transport shares the same network environment with other sub-systems and has the characteristics of large-scale data concurrency. The strategy of the CCTV to adapt to the large-scale data concurrency, the distribution model of the CCTV, and the load balancing technology during the process of distribution are introduced in this paper. Based on the performance and load measurements of the multiple servers, a load balancing mechanism is developed in view of a dynamic measurement. It has been proved by the experimental tests that this mechanism can reduce transmits delay, improve system throughput, and balance the load between the servers significantly.