Network Service Embedding in Multiple Edge Systems: Profit Maximization by Federation
Yu-Chen Tai, Li‐Hsing Yen · 2021
A service chain (SC) comprises a series of service functions realized as virtualized network functions (VNFs). Each VNF has specific resource, bandwidth, and location requirements. This study assumes multiple SCs to be placed on geo-distributed heterogeneous edge servers owned by multiple edge service providers (ESPs). Each ESP selectively hosts SCs to earn profit. We consider the problem to maximize the total profit (i.e., social welfare) of all ESPs by organizing ESPs into a set of disjoint federations called federation structure. We formulate the problem of finding an optimal federation structure, identify the difficulties in solving the problem, and present two time-efficient heuristics as our approach. Simulation results show that elaborate federation structures do bring in higher social welfare compared with simple all-in or all-out ESP organization. The proposed heuristics can also approximate the optimal result in many cases.