Energy saving in 5G substations using edge caching for adaptive immersive media streaming

As the number of wireless network-dependent devices exponentially grows, wireless access networks require capacity expansion to meet the high bandwidth demands of the applications running on these devices. This results in high power consumption, especially for services like traditional video and immersive media streaming which necessitate stringent high transmission rates. In this study, a part-based media edge caching scheme is proposed for energy-efficient wireless access networks. Intelligent-edge-assisted networks store only parts of the media content on edge servers and deliver the remaining parts when active users continually engage the streaming session. This is done by responding to network queries seeking to continue with content delivery or terminate the streaming session. The reward for a timely response is an assurance of uninterrupted streaming sessions. In this way, a streaming session without a user's response is halted to save the transmission energy and bandwidth of the edge servers. We formulate the edge caching problem as a multi-constrained multiple-knapsack problem and attempt to optimally place the content with the help of the proposed algorithm. Extensive simulations reveal that the proposed caching scheme shows improvements in the cache hit ratios and average delays as compared with LFU, FIFO, and LRU.