Title
Prefetch and Cache Replacement Based on Thompson Sampling for Satellite IoT Network
Abstract
In recent years, separating locators and identifiers has been widely applied in satellite Internet of things (IoT) networks. The identifier is the terminal identity, and the locator is used as the identification of routing. Therefore, an enormous mapping server is applied to store the mapping information between the identifier and locator, which needs to be updated timely and effectively. In this paper, a hybrid mapping server is developed to store the mapping table items, applied in the Global and Local aggregation nodes. Local aggregation nodes could cache the mapping table items locally and accelerate the mapping query by completing the query locally. In general, due to the limited storage space of Local aggregation nodes, only a proportion of the mapping information obtained from Global aggregation nodes could be stored, making the selection of cached mapping table items essential. In our work, a prefetch and cache replacement method (PCRA) based on Thompson sampling is proposed to select the appropriate mapping table items to cache, which combines the characteristics of the large number of satellite IoT terminals and fast terminal moving speed. It is shown that, compared with the traditional cache strategies, PCRA could improve the cache hit rate by around 10%.
Year
DOI
Venue
2021
10.1109/ICC42927.2021.9500508
IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS (ICC 2021)
DocType
ISSN
Citations 
Conference
1550-3607
0
PageRank 
References 
Authors
0.34
0
4
Name
Order
Citations
PageRank
Yifan Chen15819.82
Junyu Liu27720.34
Yan Shi35410.21
Min Sheng4907111.63