On Efficient Cache Management of Cloud Radio Access Networks for 5G Mobile Networks

On Efficient Cache Management of Cloud Radio Access Networks for 5G Mobile Networks

Melody Moh (San Jose State University, USA) and Deepika Pathinga Rajendiran (San Jose State University, USA)
Copyright: © 2020 |Pages: 28
DOI: 10.4018/978-1-7998-1152-7.ch007

Abstract

Cloud radio access networks (CRAN) have been proposed for 5G technologies to provide improved scalability, flexibility, and performance for supporting rapid increase of IoT devices. This chapter designs a new efficient cache management scheme for the baseband unit (BBU) pool in CRAN. First, it adopts the exponential-decay (EXD) scheme to keep recently frequently requested records in cache and enhances it with analytical hierarchy process (AHP) to support multiple levels of mobility and QoS. The other new algorithms include a probability-based scoring scheme, a hierarchical, or tiered, approach, and enhancements to previously existing approaches. Performance evaluation shows that the new schemes offer high cache hit ratios and a reduction in network traffic as compared with other existing and classic caching mechanisms. The authors believe that this work is important in advancing 5G technology for supporting IoT services and is also useful to other cache management systems.
Chapter Preview
Top

In this section, the authors first describe background information including the basic 5G/LTE-A architecture and CRAN model. This is followed by related works in cache management.

Key Terms in this Chapter

Cache Miss: Data that is wanted is temporarily not found in the cache storage. Have to write the data from cloud in case of future access.

Latency: With cache hit, it is the amount of time (delay) it takes to access and update the data. With cache miss, it is collective time it takes to evict data from cache memory and write it to cloud memory (if needed) and write and update the incoming data to the cache and cloud.

Cloud Memory: It is a secondary storage device which keeps all the data. If there is a cache miss, the data can be accessed from cloud memory.

Cache Hit: Data that is needed is readily available for access in the small cache memory.

Network Traffic: With cache miss, there is a need to travel across the data center network from the cache to the cloud storage in order to perform necessary writes. As a result, there is a certain amount of traffic is generated on the network as a result of the cache miss.

Cloud Write: Cloud Writes are represented as an average number of writes to cloud storage per user request. It is zero if it is a cache hit. With miss, the evicted data has to written in cloud as well as incoming data also be updated in cloud. So, 2 cloud writes with cache miss.

Complete Chapter List

Search this Book:
Reset