Home / Technology Policy & Ethics / July 2019 / Offload Computation in Mobile Edge Computing with Wide-Band Support of Channel Bonding

Offload Computation in Mobile Edge Computing with Wide-Band Support of Channel Bonding

By Ali Raza, University of Engineering and Technology, Pakistan, Syed Hashim Raza Bukhari and Farhan Aadil, COMSATS University Islamabad, Pakistan

July 2019

Extensive enhancement in technology enables the use of mobile and Internet of Things (IoT) devices in the high demanding applications and services. The users’ demands for quality of service (QoS) and computation intensive applications such as augmented reality, online gaming, e-health, smart home and environmental monitoring etc. are also increasing exponentially1. Although, mobile and IoT devices have become much more powerful in the last decades and can perform a variety of tasks in an efficient manner, these are still incapable for computation intensive and time critical applications2. With the low processing capacity, IoT devices cannot produce results within a given time constraint. Even with high CPU power, executing a computational expensive task depletes their battery energy and shortens the operational lifetime. Mobile cloud computing (MCC) is one of the most promising solutions in the last decade3. It empowers the resource constrained mobile and IoT devices with elastic computing power and large storage capacity. MCC enables the realization of centralized computing and allows offload computation for CPU hungry and time sensitive tasks at the cloud server. However, intrinsically the MCC has certain serious concerns such as long transmission latency, privacy/security of user data, immense increase in Internet traffic due to billions of IoT devices in the near future4. These issues motivate the emerging a new paradigm known as mobile edge computing (MEC). MEC decouples the cloud resources into edge servers which are located near the end user, usually alongside access point (AP).

In recent years, MEC has attracted researchers’ concentration, and the trend has shifted from MCC to MEC. MEC leverages the close proximity of the edge server with mobile device and significantly reduces the transmission delay, backbone network traffic and mobile node energy2. It also exploits the principal of locality which states that most of the information produced locally is consumed locally5. MEC servers have limited computational resources and are deployed in a distributed manner4.

Figure 1.  MEC system with varying link bandwidth and corresponding time over that link for sending task input data to server and receiving back the results after processing. The wider the bandwidth, the shorter the transmission time will be.

The envisioned 5G network will greatly exploit the MEC system. It is estimated that by 2020, about 20 billion IoT devices will be connected with the Internet6. These IoT devices are connected with the MEC server via AP over the radio interface. Time sensitive and computational expensive tasks are sent to the MEC server for offload computations. This offloading can be either binary (i.e. full or no offloading in small or well integrated tasks); or partial where there is a tight dependency among various sub functions of a task. The offloading decision can be taken via either a centralized or distributed manner. In the centralized environment, user equipment sends offloading parameters such as transmission power, available channel condition, input data size, local CPU power etc. to the MEC server. It then takes decision for whether to perform offload computation or compute the process locally. In the decentralized architecture, mobile users collaborate with each other and take the offloading decision in a distributed manner. Whether there is binary or partial offloading in the centralized or decentralized architecture, it can only be useful if the sum of offload computation latency is less than local computation1. Offload computations include queuing, scheduling and processing at server plus transmission time for sending input data to the server and sending back results to the user equipment. Let a task ω = dC where, d is the size of input date, C is the number of cycles required to compute 1 bit of input data.

Where fm is the mobile CPU capacity in cycles per second, B is available channel bandwidth, and Ts is task execution time at edge server. Equation (1) implies that offloading is only beneficial if the task size is large and network conditions are favorable, i.e. high bandwidth is available. If the task size is quite small with enough fm and/or poor network performance, it is better to execute the task locally at the mobile device. Furthermore, there is a direct relationship between CPU cycles utilization and energy consumption. The state-of-the-art mobile CPU architectures use dynamic voltage and frequency scaling which allows the dynamically stepping up-down of voltage with CPU utilization, resulting in an increase or decrease in energy consumption. For time critical applications, short computation latency can be achieved at the cost of increased energy consumption. Therefore, for the energy constrained devices, offloading can be leveraged if it also satisfies the following condition:


Etran is the energy consumption of a mobile node for packet transmission and Eidle is the energy consumption at the idle state when the task is offloaded. Equation (1) and (2) are jointly used to make the offloading decision with the core objectives of minimizing computation latency and optimizing energy consumption.

Alongside the improvements in the physical architecture of the mobile and IoT devices to improve the CPU power and battery capacity, there is a need for techniques which can allow the use of IoTs in computation and communication hungry applications7. For instance, the above stated objectives can also be achieved by providing favorable network conditions. By favorable network, we mean there is sufficient bandwidth available such that desired data rate can be achieved with the given Pt. In the given context of MEC, wide bandwidth can be beneficent in two folds. First, it minimizes the transmission latency and efficiently delivers the input data/output results from/to the mobile device, hence resulting in a reduced Toff.

Channel bonding (CB) is a technique which combines the set of contiguous channels to form a single channel of wide bandwidth. In the recent past, CB has been widely used in area of cognitive radio sensors network (CRSN), cognitive radio ad-hoc network (CRAHN), wireless local area network (WLAN) and cellular network8. With the support of cognitive radio (CR), CB opportunistically utilizes the spectrum resources and makes a bond of those channels having relatively very low or no traffic9, 10, combining the multiple channels resulting in a single wide-band channel. Whenever a wireless node requires sending data (especially a huge amount of data), a bond of channels is formed, and once the data transfer is completed, the bond is broken to release the channels.  Figure 1 illustrates the benefits of CB. Increasing the number of channels in the bond will reduce the data transmission time. Use of channel bonding in MEC can greatly enhance the spectral efficiency, enlarge the available bandwidth, and hence reduce the transmission latency for offload computation.

In addition to this, CB can also play the role for energy optimization. Since , CB reduces the energy consumption in two ways. First, according to the ShannonHartley theorem for channel capacity (i.e., ), where the signal-to-noise ratio is directly related with Pt, an increase in bandwidth B allows the reduction in Pt. Second, for transmitting the given number of bits d, increasing the bandwidth ultimately lowers the per bit energy consumption, and hence a large amount of data can be sent with the given energy of the wireless node.

However, there are certain issues that must be resolved before successful implementation of CB in MEC. First of all and the most important, many of the IoT devices perform simple input/output tasks and therefore, have simple architecture, whereas, CB requires complicated structure for supporting wide band channels11. Secondly, CB increases the channel width which causes harmful interference to adjacent channels, known as adjacent channel interference (ACI). Guard bands are placed between adjacent channels to minimize ACI. In order to remove ACI effectively, determining the width of guard band is an important task. Furthermore, various IoT applications have diverse sets of requirements and therefore demand for different channel bond size (i.e. bandwidth). Allocating bandwidth less or more than the application requirement will ultimately misutilize the spectral resources.


  1. Pavel Mach and Zdenek Becvar. Mobile edge computing: A survey on architecture and computation offloading. IEEE Communications Surveys & Tutorials, 19(3):1628-1656, 2017.
  2. Yuyi Mao, Changsheng You, Jun Zhang, Kaibin Huang, and Khaled B Letaief. A survey on mobile edge computing: The communication perspective. IEEE Communications Surveys & Tutorials, 19(4):2322-2358, 2017.
  3. Jasenka Dizdarević, Francisco Carpio, Admela Jukan, and Xavi Masip-Bruin. A survey of communication protocols for internet of things and related challenges of fog and cloud computing integration. ACM Computing Surveys (CSUR), 51(6):116, 2019.
  4. Ana Juan Ferrer, Joan Manuel Marques, and Josep Jorba. Towards the decentralised cloud: Survey on approaches and challenges for mobile, ad hoc, and edge computing. ACM Computing Surveys (CSUR), 51(6):111, 2019.
  5. Peter J Denning. The locality principle. In Communication Networks And Computer Systems: A Tribute to Professor Erol Gelenbe, pages 43-67. World Scientic, 2006.
  6. Ibrar Yaqoob, Ejaz Ahmed, Ibrahim Abaker Targio Hashem, Abdelmuttlib Ibrahim Abdalla Ahmed, Abdullah Gani, Muhammad Imran, and Mohsen Guizani. Internet of things architecture: Recent advances, taxonomy, requirements, and open challenges. IEEE wireless communications, 24(3):10-16, 2017.
  7. Abbas M Al-Ghaili, Hairoladenan Kasim, Marini Othman, and Zainuddin Hassan. A review on building energy efficiency techniques. International Journal of Engineering and Technology (UAE), 7(4):35-40, 2018.
  8. Syed Hashim Raza Bukhari, Mubashir Husain Rehmani, and Sajid Siraj. A survey of channel bonding for wireless networks and guidelines of channel bonding for futuristic cognitive radio sensor networks. IEEE Communications Surveys & Tutorials, 18(2):924-948, 2015.
  9. SHR Bukhari, S Siraj, and MH Rehmani. Wireless sensor networks in smart cities: Applications of channel bonding to meet data communication requirements. 2018.
  10. Mubashir Husain Rehmani. Cognitive Radio Sensor Networks: Applications, Architectures, and Challenges: Applications, Architectures, and Challenges. IGI Global, 2014.
  11. Texas Instruments. Wlan channel bonding: Causing greater problems than it solves. White Paper, 2003.

Ali Raza received the MCS degree in Computer Science from Virtual University, Lahore, and MS in Computer Science from COMSATS Institute of Information Technology, Attock, Pakistan in 2013 and 2018 respectively. Currently he is pursuing his PhD from University of Engineering and Technology Texila, Pakistan. His current research interests include Cognitive Radio networks, Flying ad hoc Networks, bio-informatics, and bio-inspired algorithms.


Syed Hashim Raza Bukhari received his Ph.D degree in Electrical Engineering in 2017from COMSATS University Islamabad (CUI), Wah Cantt, Pakistan. Earlier, he received his M.S and B.Eng. degree in Computer Engineering in 2011 and 2007 respectively. Hashim has more than 12 years of experience in academics and has received numerous appreciations upon his contributions for the improve- ment of standards. He has also received the research productivity award for his research contributions in 2017 from COMSATS University Islamabad. His research interests include the issues in wireless sensor networks with dynamic spectrum access, cognitive radio networks and ad hoc networks. He is currently editor of IEEE Future Directions Newsletter: Technology, Policy & Ethics and guest editor of Springer Journal of Network and Systems Management (JNSM). He is also reviewer of several prestigious journals including IEEE Communication Magazine, IEEE Transactions on Industrial Informatics, IEEE Transactions on Wireless Communications, IEEE Transactions on Vehicular Technology, IEEE Communication Letters, IEEE Access journal, Elseviers Computers and Electrical Engineering (CAEE) journal, Elsevier Journal of Network and Computer Applications (JNCA), Ad Hoc Sensor Wireless Networks (AHSWN) Journal, Springer Wireless Networks Journal, Elsevier Pervasive and Mobile Computing journal, and the Journal of Communications and Networks (JCN).

Farhan Aadil received his B.S. degree in Computer Science from Allama Iqbal Open University, Pakistan in 2005. He pursued a career in the computer science for 4 years (2005 to 2009). He received his M.S. & Ph.D degrees in Software Engineering and Computer Engineering in 2011 and 2016 respectively, from University of Engineering and Technology, Taxila, Pakistan. His research interests include Vehicular ad hoc Networks, Machine Learning, and Evolutionary algorithms.



Shagufta Henna is a postdoctoral researcher with the telecommunication software and systems group, Waterford Institute of Technology, Waterford, Ireland. She received her doctoral degree in Computer Science from the University of Leicester, UK in 2013. She is an Associate Editor for IEEE Access, EURASIP Journal on Wireless Communications and Networking, IEEE Future Directions, and Human-centric Computing and Information Sciences, Springer. Her current research interests include deep learning, edge intelligence, network security, machine learning for 5G and beyond, and Intent-based networking.