一种多区域多目标优化的应急广播流媒体直播算法

A Live Media Streaming Algorithm in Emergency Broadcast based on Multi-region and Multi-Target Optimization

针对当前应急广播采用短信转发、话筒语音等直播方式存在操作灵敏度差、使用不便、延迟长等问题,建立手机与广播终端之间的流媒体直播优化模型,提出一种多区域多目标优化的应急广播流媒体直播算法。该算法采用数据分片在节点的缓存时间进行等间隔划分,利用人员级别与等待时间为权重作为马尔可夫链转移概率参数,竞争下一时刻分发任务窗口,确定终端执行广播顺序。经过数据仿真测试验证了该算法在最优情况下,响应时间控制在10 s内,播放延迟时间是其它算法的1/4,在带宽大于20 M情况下,可保证500多个直播终端稳定广播,能满足应急广播的需求。

In view of the fact that the current emergency broadcast uses SMS forwarding,microphone voice and other live broadcast methods,such as poor operational sensitivity,inconvenient use,long delay,etc.,a streaming media live broadcast optimization model between mobile phone and broadcast terminal is established,and a multi-region multi-objective optimization algorithm for emergency broadcast streaming media live broadcast is proposed.The algorithm divides data fragments into equal intervals in the cache time of the nodes,and uses the weights of personnel level and waiting time as the transfer probability parameters of Markov to compete for the next time to distribute the task window to determine the broadcast order of the terminal.The simulation results show that under the optimal conditions,the response time is controlled within 10 seconds,and the playback delay time is 1/4 of the other algorithms.When the bandwidth is more than 20M,the algorithm can ensure stable broadcasting of more than 500 live terminals and meet the needs of emergency broadcasting.