Abstract
With the emergence of smart mobile devices (SMDs) and mobile applications, the cloud-mobile edge computing (MEC) collaborative computation offloading (CMCCO) scheme, i.e., offloading the computation-intensive task from the local SMD to either the MEC server, or the remote mobile cloud computing server (MCC), is widely identified as a promising candidate under the conflict between SMDs’ limited computing ability and computing-intensive application requiring higher energy consumption. Meanwhile, the existing CMCCO scenario over integrated cloud-MEC Fiber Wireless broadband access networks (CM-FiWi) architecture, by generally fixing computing ability and transmitting power, still achieves higher computation offloading overhead in terms of task’s aggregate response time and SMD’s energy consumption. In light of this, the energy-aware collaborative computation offloading (EA-CCO) paradigm with very diverse types of computation tasks over CM-FiWi broadband access network is provided in this paper. An iterative searching algorithm for collaborative computation offloading scheme (ISA-CCO) is proposed as a solution to obtain minimized task offloading overhead, which jointly takes scaling computing ability, variable transmit power, and residual battery rate into considerations. Extensive numerical results demonstrate that the proposed solution outperforms the traditional paradigms, e.g., optimal enumeration collaborative computation offloading scheme (OECCO), approximation collaborative computation offloading algorithm (ACCO), and game theoretic collaborative computation offloading scheme (GT-CCO). More specially, the proposed ISA-CCO scheme obviously achieves lower overall task offloading overhead than those fixed transmit power and computational frequency scaling.
Highlights
As moving towards the proliferation of 5G and beyond, Internet of Things (IoTs), and Tactile Internet, the coexistence of conventional broadband traffic and emerging computation offloading task poses significant challenges for communication paradigm, key enabling technologies, and operation managements [1]–[3]
From the standpoint of the current mobile edge computing (MEC) hosted networks architecture, it is known that the traditional MEC networks are the networking technology combining cellular network and core network (CN), which can be generally classified into three deployment scenarios, i.e., MEC scheme merged with radio access network (RAN) based on 4G Evolved packet core (EPC), MEC scheme integrated with CN based on 4G EPC, and MEC server deployment scheme based on 5G
Toward to deal with this issue, we introduce scaling computational frequency and variable transmit power, while taking the residual battery rate into consideration and achieve the tasks’ optimal task offloading set in cloud-MEC Fiber Wireless broadband access networks (CM-Fiber Wireless access networks (FiWi)) broadband access network
Summary
As moving towards the proliferation of 5G and beyond, Internet of Things (IoTs), and Tactile Internet, the coexistence of conventional broadband traffic and emerging computation offloading task poses significant challenges for communication paradigm, key enabling technologies, and operation managements [1]–[3]. In addition,the efficient cost optimization scheme for hybrid cloudlet placement over FiWi can be further bridge longer access delay and low latency application Toward this end, an energy-aware collaborative computation offloading problem (EA-CCO) with very diverse types of computation tasks over CM-FiWi broadband access network is formed to further save energy. An novel energy-aware three-level collaborative computation offloading scheme with very diverse types of computation tasks is considered over integrated CM-FiWi access network to advocate the coexistence of centralized cloud and distributed cloudlet. NETWORK ARCHITECTURE we first will describe the state-of-the-art CM-FiWi network architecture integrating MEC server, MCC server, and FiWi access networks, which empowers the energy-aware collaborative computation offloading paradigm and supports coexistence of local SMDs, MEC server and MCC server More special, both communication model and computation offloading model are separately discussed in the proposed system model. The larger weight coefficient ωi,j is, the larger probability SMD contents the channel resource
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
