Table 1. Summary of existing cache placement schemes.

Ref	Cache placement method	Objectives	Limitations
Chai et al. (2012)	All the intermediate nodes cache content	Align demanded content with requesting users	Data duplication, limited content diversity
Cho et al. (2012)	Cache chunks by frequency	Stored chunks by related distance	Single-factor popularity prediction reduces cache hit ratio
Suksomboon et al. (2013)	Cache by probability metric	Assign [0,1] probability to all named packets	Static threshold caused lower hit rate
Nour et al. (2020)	Place content by request count	Get popular content closer to users	Greater distance results in more hops
Bernardini, Silverston & Festor (2013)	Cache data with popularity metric	Establish and maintain local popularity tables within nodes	Single-factor estimates achieved suboptimal cache performance
Ong et al. (2014)	Cache popular content with flexible threshold	Identify popular placements through topology features	Delays due to router’s low user proximity
Wang et al. (2014)	Cache content based on popularity-routers correlation	Identify important routers based on their distribution power	Placement in centrality nodes increases hop-count
Xiaoqiang, Min & Muqing (2016)	Central nodes caching based on discrete request rates	Select contents based on their discrete arrival rates	Ignoring historical data led to low cache hit rate
Amadeo et al. (2022)	Caching on closeness-aware nodes	Create hop-count based closeness metric for content providers and consumers	Computationally expensive due to updates at forwarding nodes
Amadeo et al. (2020)	Cache content by popularity and freshness	Base caching decisions on a combination of popularity and freshness	High cache miss rate due to fixed freshness threshold
Liu et al. (2017b)	Popularity prediction in SDN delivery network using auto-encoders	Forecast popular content from request arrivals	No method defined to select optimal placements
Asmat et al. (2020)	Cache frequently accessed content with in-band communication	Nodes keep local and network popularity tables	The computational overhead increases due to dual popularity table functionality.
Zha et al. (2022)	Dynamic threshold for popular content caching	Select content using exponential weighted average	Limited history-based selection led to low cache hit
Liu et al. (2019)	Autoregressive model-driven popular content placement in NDN devices	Maintained the lagged values of relevant popularity variables	High retrieval delays due to absence of strategical placement scheme.