scispace - formally typeset
Search or ask a question
Journal ArticleDOI

Mobile Edge Computing: A Survey

01 Feb 2018-IEEE Internet of Things Journal (IEEE)-Vol. 5, Iss: 1, pp 450-465
TL;DR: The definition of MEC, its advantages, architectures, and application areas are provided; where the security and privacy issues and related existing solutions are also discussed.
Abstract: Mobile edge computing (MEC) is an emergent architecture where cloud computing services are extended to the edge of networks leveraging mobile base stations. As a promising edge technology, it can be applied to mobile, wireless, and wireline scenarios, using software and hardware platforms, located at the network edge in the vicinity of end-users. MEC provides seamless integration of multiple application service providers and vendors toward mobile subscribers, enterprises, and other vertical segments. It is an important component in the 5G architecture which supports variety of innovative applications and services where ultralow latency is required. This paper is aimed to present a comprehensive survey of relevant research and technological developments in the area of MEC. It provides the definition of MEC, its advantages, architectures, and application areas; where we in particular highlight related research and future directions. Finally, security and privacy issues and related existing solutions are also discussed.

Summary (13 min read)

Jump to: [Introduction][1.2 Thesis Outline][2.1 Mobile Edge Computing][2.1.1 Definition of Mobile Edge Computing][2.1.2 Related Concepts and Technologies][History and Role of RAN in Cellular Networks][Three-Layer Architecture][Adaptive Computation Offloading][2.1.4 Advantages of Mobile Edge Computing][2.2.1 Network Functions Virtualization][2.2.2 Software-Defined Network (SDN)][2.2.3 Fifth Generation Wireless Networks][2.3.1 Live Video Streaming][2.3.2 Internet of Things (IoT)][2.4 Related Surveys][3.1 Applications][3.1.1 Augmented Reality (AR)][3.1.2 Content Delivery and Caching][3.2 Emerging Scenarios][3.2.1 Healthcare][3.2.2 Mobile Big Data Analytics][3.2.3 Connected Vehicle][3.2.4 Video Analytics][3.2.5 Smart Grid][3.2.6 Wireless Sensor and Actuator Networks (WSAN)][3.2.7 Smart Building Control][3.2.8 Ocean Monitoring][3.3.1 Low Latency][3.3.2 Computational Offloading][3.3.3 Storage][3.3.4 Energy Efficiency][4.1 MEC Platform][4.2 Deployment Scenario][4.3.1 5th generation test network][4.3.2 Industrial Testbeds][5.1 Security][5.1.1 CIA Triad][2. Integrity:][5.1.2 Network Security][5.1.3 Core Network Security][5.1.4 MEC Server Security][5.1.5 Virtualization Security][5.1.6 End Devices Security][5.2 Privacy Issues][5.3 Security Mechanisms][5.3.1 Identification and Authentication][5.3.2 Access Control][5.3.3 Network Security Mechanism][5.3.4 Virtualization Security Mechanism][5.3.5 Data Security][5.3.6 Data Computation Security][Open Research Problems][6.1 Security][6.2 Resource Optimization][6.3 Transparent Application Migration][6.4 Pricing][6.5 Web Interface][6.6 Other Issues][7.1 The problem statement][7.2 Challenges during the project][7.2.1 Available testbeds][7.2.2 Benchmark][7.2.3 Available applications][7.3 Thesis Contributions] and [Chapter 8 Conclusion]

Introduction

  • The prevalence of mobile terminals, such as a smartphones or tablet computers, has an uttermost effect on mobile and wireless networks that has triggered challenges for mobile networks worldwide [20] [92] .
  • Cellular networks has to endure low storage capacity, high energy consumption, low bandwidth and high latency [68] .
  • Mobile cloud computing (MCC) that is an integration of cloud computing to mobile environment has provided considerable capabilities to the mobile devices that empowers them with storage, computation and energy by proffering the centralized cloud resources [59] [25] .
  • The trend of increase in mobile usage is fundamentally driven by the augmentation of mobile users and mobile application development (e.g., iPhone apps, Google apps etc.) [45] [13].
  • MEC offers MCC capabilities by deploying cloud resources, e.g., storage and processing capacity, to the edge within the radio access network that leverage end user with swift and powerful computing, energy efficiency, storage capacity, mobility, location and context awareness support [100] [41] .

1.2 Thesis Outline

  • This thesis presents a survey on mobile edge computing that is organized in the following way: Chapter 1 gives a brief introduction of mobile edge computing and its value in the mobile operator networks.
  • Chapter 2 (Background and Related Surveys) describes an overview of mobile edge computing that mainly encompass; definition, architecture, mobile edge computing advantages key enablers and related surveys that are presented recently.
  • Security, resource optimization, tranparent application migration, pricing, web interface and other issues are briefly discussed.
  • Some of the future works are also stated.

2.1 Mobile Edge Computing

  • The term 'mobile edge computing' was first introduced in 2013 when Nokia Siemens Networks and IBM developed MEC platform that enable applications to run directly.
  • This platform accelerates only the local scope that does not support application migration, interoperability etc. [77] .
  • Later, in 2014, MEC was standardized by European Telecommunications Standards Institute (ETSI) Industry Specification Group (ISG), the group includes Nokia Networks, Intel, Vodafone, IBM, Huawei and NTT DOCOMO.

2.1.1 Definition of Mobile Edge Computing

  • According to European Telecommunications Standards Institute (ETSI), mobile edge computing is defined as [36] : "Mobile Edge Computing provides an IT service environment and cloud-computing capabilities at the edge of the mobile network, within the Radio Access Network (RAN) and in close proximity to mobile subscribers.".
  • Being deployed at a nearest location, mobile edge computing has an advantage to analyze and materialize big data, also known as 2. Proximity.
  • It is also beneficial for compute-hungry devices, such as augmented reality, video analytics etc.
  • MEC receives information from edge devices within the local access network to discover device location.

History and Role of RAN in Cellular Networks

  • Back in early 1980s, first commercial cellular network (1G generation) was introduced with the compliance of analog modulation and mobility support, which later was eventually replaced by 2G because of its digital radio signaling capability using time division multiple access (TDMA).
  • With an accustomed support of mobile internet using RAN Long-Term Evolution (LTE), 4G got an edge over other wireless mobile telecommunications technology providing best user experience [44] .
  • In traditional cellular radio system, wireless user equipments connects through RAN to the mobile operator networks.
  • The RNCs are connected with one or two back haul networks.
  • The UMTS is a third generation system that may depend on global system for mobile communication (GSM) that has been developed in Europe.

Three-Layer Architecture

  • MEC is a layer that resides between cloud and mobile devices.
  • Mobile edge computing mostly complies with cloud computing to support and enhance performance of the end devices.
  • These servers are implemented locally at mobile user premises, e.g., parks, bus terminals, shopping centers, etc. [56] .
  • MEC can be deployed at LTE base station or multi-technology (3G/LTE) cell aggregation site.
  • The multi-technology cell aggregation site can be both indoor or outdoor location.

Adaptive Computation Offloading

  • Computation offloading is a solution to enhance the capacity of mobile devices by transferring computation to higher resourceful servers that are located at the external location [50] .
  • Emergence of resource-demanding applications, such as 3D games will continue to demand more mobile resources.
  • Improvement of mobile hardware and network will still not be able to cope up with the trend in demand.
  • Therefore, mobile devices will always have to compromise with its limited resources, such as resourcepoor hardware, insecure connection and energy driven computing tasks [47] .
  • To deal with these constraints, many researchers have managed computation offloading to computational power resources [33] [82] [49] , such as cloud.

2.1.4 Advantages of Mobile Edge Computing

  • As already discussed in previous sections, there are several benefits associated with mobile edge computing that is turning out to be promising for both mobile network operators (MNOs), and application service provider (ASP), in addition also befitting content providers, Over-the-top (OTT) players, network equipment vendors, IT and middleware providers [107] [13] .
  • Moreover, power consumption is also one of the main concerns.
  • In addition, distributed virtual servers provision scalability and reliability.
  • Application service providers could gain profit by MEC enabled infrastructure-as-a-service (IaaS) platform at the network edge that make ASPs services scalable along with high bandwidth and low latency.
  • ASPs could also get a real time access to the radio activity that may develop more capable applications.

2.2.1 Network Functions Virtualization

  • Network functions virtualization enables the virtualize environment of network services that are launched by the dedicated hardware.
  • The goal of NFV is to move network functions from dedicated hardware devices to generic servers.
  • NFV comes with several beneficial attributes, such as flexibility, cost effectiveness, scalability and security.
  • According to the change in demands, NFV enables a flexible access to the operators and service providers to scale there services.
  • Virtualize network devices installed at the network edge will be beneficial to end users by integrating MEC in the virtualize environment.

2.2.2 Software-Defined Network (SDN)

  • Software-defined network (SDN) is an innovation to computer networking that separates control layer and the data layer [84] .
  • Data layer contains user generated messages and is responsible to forward them using the forwarding tables prepared by the control layer [39] .
  • MEC concept along with SDN can make centralized control more efficient and reliable, e.g., in vehicle to vehicle connectivity the ratio of packet loss can be resolved.

2.2.3 Fifth Generation Wireless Networks

  • The 5th generation wireless system to be the next communication standards that are likely to be more faster and more reliable then 4G networks.
  • 5G together with MEC can possess better user experience.
  • MEC at the edge of the network will be providing services for complex traffic handling and routing.

2.3.1 Live Video Streaming

  • Live video streaming, such as live TV or live conferencing on mobiles devices requires high bandwidth and ultra low latency.
  • This data stream creates a huge traffic that stresses the mobile network.
  • Moreover, heavy data movement over the network refers to service interruption or service denial.
  • Since live video streaming is one of the main goal of 5G networks, MEC will play a major role for video streaming by pushing intelligence at the network edge near to the end user.

2.3.2 Internet of Things (IoT)

  • IoT is an emerging technology in which physical objects communicate with each other mainly through internet.
  • These physical object requires fast data transmission and high computational power in order to keep there data integrity.
  • IoT can largely benefit from MEC technology and deliver better services.

3.1 Applications

  • MEC architecture is a new revenue stream for mobile operators that yet had to get mature but on the other hand the authors see quite a few areas adopting Edge Computing (e.g Fog Computing) as it is been compassed in recent articles [36] [23] .
  • Some recognized applications include Augmented Reality, Content Delivery, Healthcare relevant applications (e.g U-Fall) etc. appears in this section.

3.1.1 Augmented Reality (AR)

  • In the era of mobile technology, augmented reality applications have recently adapted mobile technology, such as Layar, Junaio, Google Goggles, and Wikitude [67] .
  • Recently AR applications, are also being adaptive in sound and visual components, such as news, TV programs, sports, object recognition, games etc. [103] .
  • AR systems usually demand high computing power; to perform computational offloading, low latency for better quality of experience (QoE) and high bandwidth that is conducive to sustain interminable IT services.
  • Therefore, offloading computation-intensive operations at the nearest cloudlet is more optimized and efficient that could enhance user experience.
  • The application serves by integrating wireless electroencephalogram (EEG) headsets, smart phones and edge server.

3.1.2 Content Delivery and Caching

  • The edge computing technology plays a comprehensive role in Web site performance optimization, such as caching HTML content, reorganizing web layout and resizing web components.
  • User makes HTTP requests that passes through the edge server.
  • The edge computing infrastructure is time efficient as compared to the traditional internet infrastructure where user requests are handled at the servers that are distantly placed at the service provider.
  • Under congested network conditions where several users are streaming video at the same time, the graphics resolution is decreased to minimal to accommodate every user averting any denial of service or jitter.
  • MEC incorporated with internet infrastructure can bring intelligence, such as website optimization within the RAN.

3.2 Emerging Scenarios

  • Emerging scenarios of MEC are demonstrated that are recently considered in the ETSI white paper [36] , such as video analytics and mobile big data.
  • Several papers [102] [97] [43] [74] have referred MEC scenarios in connected vehicle, smart grid and wireless sensor and actuator networks (WSAN).
  • Further more, [90] expanded the scenario on smart building control and software-defined network (SDN), later followed by ocean monitoring [3] .

3.2.1 Healthcare

  • Science and technology in health domain is a substantial research area for many researchers [19] .
  • Falls are common among stroke patients who suffers mostly due to hypoglycemia, hypotension, muscle weakness, etc.
  • Recently, researchers have proposed smart healthcare infrastructure called U-Fall, that exploit smartphones by engaging edge computing technology.
  • U-fall sense motion detection with the help of smart device sensors, such as gyroscopes and accelerometers.
  • In addition, the proposed infrastructure is capable to deliver accurate results that makes it more reliable and dependable.

3.2.2 Mobile Big Data Analytics

  • Mobile phone technology is valued a growth-engine for small, medium and large enterprises, and also have widespread social connotation.
  • The ubiquity of mobile phones and its big data coming from applications and sensors, such as GPS, accelerometer, gyroscope, microphone, camera and bluetooth are stressing the network bandwidth [52] .
  • Big data analytics is a process of extracting meaningful information from raw data that could be helpful for marketing and targeted advertising, customer relations, business intelligence, context-aware computing, health care etc. [8] [78] .
  • Instead of using typical path from edge device to the core network, big data can be collected and analyzed at the nearest MEC location.
  • This scenario will perhaps also accommodate data coming from several IoT devices for big data analytics.

3.2.3 Connected Vehicle

  • Vehicles are facilitated with an internet access that allows them to connect with other vehicles on the road.
  • The connection scenario can either be vehicle-to-vehicle, vehicle to access point or access point to access point.
  • By deploying MEC environment along side the road can enable two-way communication between the moving vehicles.
  • One vehicle can communicate with the other approaching vehicles and inform them with any expected risk or traffic jam, presence of any pedestrian and bikers.
  • In addition, MEC enables scalable, reliable and distributed environment that is synced with the local sensors [24] .

3.2.4 Video Analytics

  • Surveillance cameras in old times use to stream data back to the main server and then the server decides how to perform data-management.
  • Due to the growing ubiquity of surveillance cameras, old client-server architecture might not be able to stream video that may be coming from million of devices and therefore, it will stress the network.
  • In addition, MEC enabled surveillance cameras can be effective for several applications, such as traffic management application on the basis of traffic patterns can detect traffic jam or an accident.
  • The application can also be helpful for face recognition, for example, if someone commits a crime then his photo can be transferred to these intelligent cameras to trace the culprit [35] [38] [35] .
  • Perhaps, the management server make decisions as per the defined rules.

3.2.5 Smart Grid

  • Smart grid infrastructure is an electrical grid that consists several components, such as smart appliances, renewable energy resources, and energy efficiency resources.
  • Smart meters that are distributed over the network are used to receive and transmit measurements of the energy consumption [58] .
  • All the information collected by smart meter is supervised in supervisory control and data acquisition systems that maintain and stabilise the power grid.
  • Moreover, MEC integrated with distributed smart meters and micro grids can support SCADA systems.
  • MEC will balance and scale the load according to the information shared by other micro grids and smart meters.

3.2.6 Wireless Sensor and Actuator Networks (WSAN)

  • Wireless sensors and actuator networks(WSAN) are sensors that is used for surveillance, tracking, and monitoring of physical or environment situation, e.g., light intensity, air pressure, temperature etc. [48] .
  • MEC enabled actuators autonomously manage measurement process by developing an active feedback loop system.
  • Air vent sensors manages air pressure flowing in and out of the mine to save miners from any emergency.
  • These sensors consume very less energy and bandwidth with the help of MEC.

3.2.7 Smart Building Control

  • Smart building control system consists of wireless sensors that are deployed in different parts of the building.
  • Sensors are responsible for monitoring and controlling building environment, such as temperature, gas level or humidity.
  • In smart building environment, sensors installed with MEC becomes capable of sharing information and become reactive to any abnormal situation.
  • These sensors can maintain building atmosphere on the basis of collective information received from other wireless nodes.
  • If humidity detected in the building then MEC can react and perform actions to increase air in the building and blow out the moisture.

3.2.8 Ocean Monitoring

  • Scientists are researching to cope with any ocean cataclysmic incidents and know the climate changes in advance.
  • This can help to react quickly and mitigate to prevent from any disastrous situation.
  • Sensors deployed at some location in the ocean transmits data in great quantity that require large computational resources [3] .
  • The data handled by cloud may occur delays in the transmission of live forecast.
  • MEC can play a vital role to prevent for any data loss or delay in sensor data.

3.3.1 Low Latency

  • MEC is one of the promising edge technologies that improves user experience by providing high bandwidth and low latency.
  • In 2016, Abdelwahab et al [1] proposed REPLISOM that is the edge cloud architecture and LTE enhance memory replication protocol to avoid latency issues.
  • By integrating MEC and 5G, it empowers real time collaboration systems by leveraging with context-aware application platform.
  • Due to the high mobility of vehicles, smart grid environment supported by MEC use to monitor large data sets transmitted by several smart devices.
  • According to the data movement, these devices makes computation charging and discharging decisions with respect to message transmission delay, response time and high throughput network for movable vehicles.

3.3.2 Computational Offloading

  • Computational offloading is one of the main advantage of MEC to improve application performance, energy consumption and response time.
  • They first tested a single user offloading computational task at cloud server where resulting problem is non-convex optimization.
  • MEC provider receives the payment from vehicles on the basis of the amount of computational task they offloaded at MEC servers.
  • FemtoCloud client computing service is installed on each mobile device to calculate device computing capability, energy information and capacity for sharing with other mobile devices.
  • Task and scheduling module assigns user devices on the basis of the information collected fomr previous modules.

3.3.3 Storage

  • User end devices with limited storage capacity may leave negative impact of user experience.
  • End users can utilize MEC storage resources to overcome their device storage limitation.
  • In 2016, Jararweh et al [41] proposed Software Defined System for Mobile Edge Computing.
  • The proposed framework connects software defined system components with MEC to further extend MCC capabilities.
  • The components jointly works cohesively to enhance MCC into the MEC services.

3.3.4 Energy Efficiency

  • As previously mentioned, MEC architecture is designed to improve energy consumption of user devices by migrating compute intensive tasks to the edge of network.
  • DroidCloud has several modules that are shown in figure 3 .7.
  • These mobile devices are enable to share both the energy and computational resources depending on their available capacity.
  • In the proposed system, encoding techniques are wisely used to stream video on MEC server.
  • In the paper, it concludes that nDCs may lead to energy savings if the applications, especially IoT applications that generate and process data with in user premises.

4.1 MEC Platform

  • The main services of MEC application server is Commercial-Off-The-Shelf (COTS) products that is available for general mobile users.
  • As shown in figure 4 .1, MEC server is comprised of an application platform and hosting environment which is further divided into virtualization and hardware resources.
  • These services are managed by application platform management.
  • Traffic offload function (TOF) is responsible for traffic offloading on the basis of the policy that is defined.
  • Radio network information services (RNIS) enable cloud application services that serves the mobile users with in the radio access network.

4.2 Deployment Scenario

  • As mentioned earlier, mobile edge computing can be deployed flexibly and intelligently at different sites that includes UMTS radio access network , LTE E- Node B, 3G Radio Network Controller (RNC) and multi-Radio Access Technology (RAT), as illustrated in figure 4.2.
  • MEC deployment will use network functions virtualization (NFV) architecture or NFV platform may be dedicated for MEC, otherwise will be shared with MEC architecture.
  • According to the first release of information services group (ISG) MEC, the implement scenarios can either be at outdoor environment, such as LTE site, 3G site etc. or indoor environment, such as shopping malls, hospitals, etc. 2. MEC in indoor scenario:.
  • Its deployment in machine-to-machine environment can monitor temperature, humidity, air conditioning, etc. with the help of connected sensors at various indoor locations.
  • MEC can also be beneficial in case of any emergency situation, such as in any hazardous situation in a residential building environment it can help people to evacuate the building with the help of AR services etc.

4.3.1 5th generation test network

  • The 5th generation test network (5GTN) architecture was developed and successfully tested at Oulu, Finland, that is based on LTE and LTE-Advanced (LTE-A) technology [75] .
  • It opens an opportunity for application developers to develop their application in a test environment before they are brought to the market.
  • The introduced testbed is composed of different environments, one is located at Technical Research Centre of Finland (VTT's) 5G laboratory and other is at the University of Oulu's Centre for wireless communications (CWC).
  • CWC network is opened for public users, whereas VTT's network is in more secured and private environment.
  • Both networks are integrated with the help of carrier-grade technology that offers a real-time environment.

4.3.2 Industrial Testbeds

  • Nokia and China mobile successfully tested advance mobile solutions for utmost mobile data capacity and real-time video [64] .
  • Platform built for MEC with airframe Radio Cloud platform for MEC and Airscale Wi-Fi with flexi zone controllers.
  • Spectators are able to see four video feeds at the same time that are on a split mobile screen.
  • The second use test case is cooperative passing assistant that also utilizes cloudlets deployed at LTE base stations.
  • Vehicles changing lanes are alarmed with the critical distance between them.

5.1 Security

  • Mobile edge computing is not a panacea.
  • There are some challenges that need extensive research studies of every layer of MEC infrastructure.
  • This section explains MEC research issues that are mentioned in several papers, in context of different architectural designs [77].

5.1.1 CIA Triad

  • The components of CIA triad, confidentiality, integrity, and availability makeup a model design for information security.
  • There are several aspects of trust that need considerations in MEC infrastructure.
  • There are several applications hosted at the edge of network providing there services to the mobile users, e.g locationawareness.
  • In spite of the fact that these applications are beneficial but they also posses confidential risks.
  • At application layer there is no rule defined to separate user identity from its geolocation [87] .

2. Integrity:

  • The MEC ecosystem incorporates multiple actors, such as end users, service providers, infrastructure providers etc. that causes several security challenges.
  • This scenario can invite several attacks, such as man-in-themiddle attack in which the attacker can authenticate themselves to the central cloud systems and later with end devices to steal there secret information.
  • Due to less isolated environment, MEC system may suffer denial of service (DOS) attacks that could be application or packet-based, also known as 3. Availability.
  • On single node, these attacks might not be much hazardous but if the correlative attacks occur simultaneously at multiple geo-locations, it can lead to serious implication.
  • Compromised sensors in industrial sector will make a ripple effect globally.

5.1.2 Network Security

  • The preponderance of various communication networks, such as mobile core networks or wireless networks, network security is a very important element in MEC environment.
  • Hacker hijacking the network stream can launch attacks to effect MEC system performance.
  • Man-in-the-middle attack is likely to be effective before compromising gateway and later intercepts data communication.
  • Attacker can successfully manipulate data traveling from cloud to user and vice versa.
  • It is difficult to mitigate such attacks because, one of the reason is deploying and dropping virtual machines that makes it cumbersome to maintain blacklist.

5.1.3 Core Network Security

  • And most of the core network security is enable by mobile core networks or central cloud.
  • Cloud service security are mostly managed by third party suppliers, such as Amazon, Microsoft, Google etc.
  • This type of security issue will not effect the whole ecosystem and will be limited due to its decenrtalised nature.
  • Core network elements that are compromised can disrupt lower level infrastructure.
  • Attackers may have all the access to the information and may tamper the network data flow.

5.1.4 MEC Server Security

  • MEC at the edge, comprises of several virtualized servers that provide IT intelligent services.
  • This particular attack is limited to a specific geographical location and may not be very critical.
  • Being newly preambled in the technology world, MEC lacks some security expertise for adequate system security.
  • Once the login has been accessed to the MEC system resources, attacker can abuse system integrity or can execute denial of service attacks, man-in-the-middle attacks etc.
  • Another security issue is the compromise of an entire data centre.

5.1.5 Virtualization Security

  • In core mobile edge data center, several network instances co-exist sharing network instances.
  • If one resources is compromised, it can effect the whole virtualized infrastructure.
  • Several APIs implemented in MEC environment are responsible to deliver information of physical and logical infrastructure.
  • These APIs are most likely to be less protected against any malevolent activity.
  • Users moving across different geographical location can escalate such attacks to other MEC virtualized servers.

5.1.6 End Devices Security

  • End user devices can potentially be harmful to effect some elements of ecosystem.
  • The impact could be narrow due to limited user device surroundings.
  • In addition, there could be rogue users that can intrude the system and do some malicious activities.
  • They can inject false values or information in the system.
  • Moreover, there are some scenarios where devices can participate in service manipulation.

5.2 Privacy Issues

  • Due to the close proximity to end user, privacy security, such as data, usage and location may be challenging in mobile edge computing.
  • User privacy breach may get worse if the attacker gains personal information, such as credit card information, personal emails etc.
  • Attacker may retrieve user information by learning usage pattern of user device while accessing MEC.
  • Non-Intrusive Load Leveling (NILL) has been introduced to encounter these kind of issues [60] .
  • MobiShare system is flexible and secure for sharing information of geo-location and has a good support for location-base applications [98] .

5.3 Security Mechanisms

  • Security breach may cause potential harmful problems within the system.
  • Therefore, it is very important to implement security mechanism and safeguard the MEC resources from any intrusion.
  • Some of the security mechanisms are listed in this section.

5.3.1 Identification and Authentication

  • In a cloud computing environment, data centers are mostly hosted by cloud service providers, whereas in all edge paradigms, providers may be hosted by several providers depending on their choices.
  • Every entity in the ecosystem, such as end devices, virtual machine services, Cloud and MEC infrastructure service providers, and application service providers should be able to identify and mutually authenticate each other.
  • A user-friendly solution has been introduced that provides a secure authentication in a local ad-hoc wireless network [86] .
  • Similarly, NFC also enables a secure authentication method in a cloudlet scenario [15] .
  • If the connection is fragile then Stand-Alone authentication would be able to authenticate users with the cloud servers.

5.3.2 Access Control

  • Without any proper access control mechanism it is difficult to prevent MEC infrastructure from any malicious activity.
  • MEC service providers deploying VMs, these VMs connecting to the APIs available at the edge of network etc., all these resources need an access control to execute their services.
  • In the context of mobile edge paradigm, where there are many actors involved, it is very important to have an authorization mechanism that authenticate these actors to enforce there own security policy.
  • If there will be a trust relationship developed, then it will able to authorize all entities to communicate with each other abiding by the security policies.

5.3.3 Network Security Mechanism

  • Network security is one of the prime concern for MEC concept, due to the predominance network infrastructure.
  • Intrusion detection and prevention mechanism is an important prerequisite before the deployment of MEC infrastructure.
  • Several network entities could be vulnerable to any threat that need to be monitored from internal or external threat.
  • A Cloudlet that is located at one hop away from mobile devices can efficiently be meshed to form a security framework to detect any intrusion [81] .
  • SDN can isolate network traffic and segregate malicious data.

5.3.4 Virtualization Security Mechanism

  • Virtualization technology is one of the foundation for edge paradigm and thus its security is of paramount importance.
  • Malicious element getting an access to the virtual servers may hijack the whole edge data center.
  • Virtualized servers and their hosted physical servers can be protected through hypervisor hardening, network abstractions, isolation policies etc. [73] .

5.3.5 Data Security

  • This invites some challenges, such as data integrity and authorization, for example, outsourced data can be modified or disappeared, moreover, the uploaded data can easily be accessed by malicious activists.
  • Moreover, data owners and data servers posses dissimilar identities and business interests that makes MEC architecture more vulnerable.
  • Good auditing methods can be used to audit the data storage in order to confirm that data is properly stored in the cloud [99] .

5.3.6 Data Computation Security

  • Secure data computation is another important issue that has to be addressed carefully.
  • Verifiable computing allows the computing node to offload some functions to other servers that could not be trusted, but it enables the maintenance of the results that are verifiable.
  • One of the popular security services is a keyword search that means to search keywords from the encrypted data files.
  • The index is secured through one-to-many order-preserving mapping approach.

Open Research Problems

  • As a recent technology approach, no sufficient research study has been specifically designed for mobile edge computing.
  • Therefore, there are some security issues in MEC that need to be addressed before its commercial deployment.
  • This chapter illustrates and identifies the open issues that investigated by different researchers in the development of MEC.

6.1 Security

  • It is one of the main concern for technology advisers to secure MEC deployment.
  • There are some security mechanisms that are applicable in MEC, as discussed in previous chapter.
  • There are still some issues that need proper research study.
  • Moreover, different users connected to common physical server also raise some security issues [88] .
  • The application data movement is possible through encryption and decryption strategy but it effects application performance.

6.2 Resource Optimization

  • Promoting cloud infrastructure to the network edge, MEC incorporate less resources then the traditional cloud infrastructure.
  • Computational offloading is performed at the MEC virtualized servers.
  • Computational tasks carry extra overload due to heterogeneous processor architecture, for example, smart phones and cloud has mostly ARM and x86 architecture, therefore they need to perform translation or emulation [85] .
  • Thus, an optimized solution for enhancing performance of intrinsic limited resources is required [4] .

6.3 Transparent Application Migration

  • As mentioned previously, user applications tasks are moved to MEC server for computation etc..
  • It is very challenging to transparently migrate these applications for the usability of delay-sensitive mobile applications, such as real-time applications [7] .
  • Poor Compute resources and service delay, deteriorates the performance of mobile applications [5] .
  • Application migration is a software level solution that can be achieved by doing more research study to find solutions that optimize cloud services at the edge [6] .

6.4 Pricing

  • Mobile edge computing environment involves several actors that quote different prices for their services.
  • These actors have different payment methods, different customer management and different business policies.
  • Game application on user device have to utilize cloud resource, mobile network and game services.
  • The user has to pay for the game that has to be divided equally or as per mutual contract to all the entities involved.
  • This can be argued that agreeing to the pricing may be difficult among different entities.

6.5 Web Interface

  • Currently, the interface available to access MEC and cloud is the web interface that is not sufficient due to its overhead problem.
  • The web interface is generally not designed for mobiles and hence have compatibility issues.
  • Therefore, the standard protocol is required for smooth communication between the user, MEC and cloud.
  • The latest version of HTML5 is designed specifically for advanced devices, such as mobile or smart phones.
  • A performance and test based research is required to accept HTML5 in MEC environment.

6.6 Other Issues

  • Many issues that are already discussed in previous sections, in addition, there are also some other issues that are imperative to strengthen the MEC framework.
  • A comprehensive scientific research study is required to avoid any security issue that can damage the system.
  • Openness of the Network: Mobile core network has a sound authority over the mobile network but in MEC architecture it will be very challenging to open network for third party vendors due to the possible security risks.
  • This scenario causes complexity for seamless third party services.
  • Deploying resources at MEC is very important to enable robustness of the MEC server, also known as Robustness and Resilience.

7.1 The problem statement

  • The prime objective of this project is to address the research problems, as defined earlier in section 1.1.
  • What are the main challenges in using mobile edge computing and what are the solutions related to these challenges.
  • To adequately address the problem statement, MEC approaches and implementation is thoroughly studied in this thesis.
  • Available edge paradigms are extensively studied and their differences are also depicted in this project.
  • In order to address these challenges some possible solutions are identified.

7.2 Challenges during the project

  • During the project period there has been a lot of challenges in finding the relevant research study.
  • One obvious reason is that there is very few research study that specifically explain MEC concept and its implementation.
  • Therefore, there is not much scientific research conducted as yet.

7.2.1 Available testbeds

  • It was very challenging to refer any test scenario that is associated with MEC, except the testbeds mentioned in Section 4.2.
  • Those which are mentioned, does not specify the sufficient detail of the system environment, such as the system specification and components that are used during the test.

7.2.2 Benchmark

  • No specific test were found during the research study that establishes a benchmark.
  • Therefore, there is not much study available on MEC that may evaluate its performance.

7.2.3 Available applications

  • Since, MEC is not practically deployed in the industry, thus its a challenge to find MEC applications that are tested and analyzed.
  • Instead, applications that belongs to other similar edge computing technology were mentioned in this project.

7.3 Thesis Contributions

  • This project demonstrates a detail survey on MEC.
  • The findings of this thesis indicates the importance of MEC for computational intensive applications that require high bandwidth and highly latency-intolerant.
  • Furthermore, MEC longer-term role in evolved mobile networks is also indicated in this thesis.
  • The difference between the similar technologies, such as cloudlet, local cloud, fog computing and mobile cloud computing, are also demonstrated.

Chapter 8 Conclusion

  • MEC trends to provide elastic resources at the edge of the networks to the applications that are compute-intensive and demand high bandwidth and ultra low latency, especially in the scope of 5G networks.
  • MEC deployment can build an ecosystem involving third-party partners, content providers, application developers, OTT players, network vendors and multiple network operators.

Did you find this useful? Give us your feedback

Content maybe subject to copyright    Report

Mobile Edge Computing: A
Survey
architecture, applications, approaches and challenges
Nasir Abbas
Master’s Thesis Autumn 2016


Mobile Edge Computing: A Survey
Nasir Abbas
December 12, 2016

ii

Abstract
Mobile edge computing (MEC) is an emergent architecture where cloud
computing services are extended to the edge of networks into the mobile
base stations. As a promising edge technology, it can be applied to mobile,
wireless and wireline scenarios, using software and hardware platforms,
located at the network edge in the vicinity of end users. MEC provides
seamless integration of multiple application service providers and vendors
towards mobile subscribers, enterprises and other vertical segments. It is
an important component in the proposed 5G architecture that supports
variety of innovative applications and services where ultra low latency
is required. However, there are some challenges exists in the MEC eco
system. To address these challenges, first off need to understand the
network infrastructure of MEC, cloud and cellular network.
Some questions and problems are addressed in this thesis that outlines
the importance and challenges of MEC deployment. Impact of MEC
integration with the traditional mobile and cloud network appears in
this paper. A survey has been presented that contributes in general
understanding of mobile edge computing (MEC). Readers will have
an overview of MEC, such as definition, advantages, architectures and
applications. Moreover, related research and future directions are pointed
out in this thesis. Finally, security and privacy issues and their possible
solutions are also discussed.
iii

Citations
More filters
Journal ArticleDOI
TL;DR: The concept of federated learning (FL) as mentioned in this paperederated learning has been proposed to enable collaborative training of an ML model and also enable DL for mobile edge network optimization in large-scale and complex mobile edge networks, where heterogeneous devices with varying constraints are involved.
Abstract: In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications, e.g., for medical purposes and in vehicular networks. Traditional cloud-based Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL.

895 citations

Journal ArticleDOI
TL;DR: A detailed review of the security-related challenges and sources of threat in the IoT applications is presented and four different technologies, blockchain, fog computing, edge computing, and machine learning, to increase the level of security in IoT are discussed.
Abstract: The Internet of Things (IoT) is the next era of communication. Using the IoT, physical objects can be empowered to create, receive, and exchange data in a seamless manner. Various IoT applications focus on automating different tasks and are trying to empower the inanimate physical objects to act without any human intervention. The existing and upcoming IoT applications are highly promising to increase the level of comfort, efficiency, and automation for the users. To be able to implement such a world in an ever-growing fashion requires high security, privacy, authentication, and recovery from attacks. In this regard, it is imperative to make the required changes in the architecture of the IoT applications for achieving end-to-end secure IoT environments. In this paper, a detailed review of the security-related challenges and sources of threat in the IoT applications is presented. After discussing the security issues, various emerging and existing technologies focused on achieving a high degree of trust in the IoT applications are discussed. Four different technologies, blockchain, fog computing, edge computing, and machine learning, to increase the level of security in IoT are discussed.

800 citations


Cites background from "Mobile Edge Computing: A Survey"

  • ...can be done at the edge nodes and only the summarized data, if required, needs to be sent to the cloud [190]....

    [...]

Journal ArticleDOI
TL;DR: This paper provides a tutorial on fog computing and its related computing paradigms, including their similarities and differences, and provides a taxonomy of research topics in fog computing.

783 citations


Cites background from "Mobile Edge Computing: A Survey"

  • ...A recent survey on MEC can be found in [81]....

    [...]

Posted Content
TL;DR: In a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved, this raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale.
Abstract: In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications. Traditional cloudbased Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL

701 citations


Cites background from "Mobile Edge Computing: A Survey"

  • ...For example, the authors in [19], [38], [39] and [40] discuss strategies for optimizing computation offloading for mobile edge networks, but do not...

    [...]

Journal ArticleDOI
TL;DR: An in-depth survey of BCoT is presented and the insights of this new paradigm are discussed and the open research directions in this promising area are outlined.
Abstract: Internet of Things (IoT) is reshaping the incumbent industry to smart industry featured with data-driven decision-making. However, intrinsic features of IoT result in a number of challenges, such as decentralization, poor interoperability, privacy, and security vulnerabilities. Blockchain technology brings the opportunities in addressing the challenges of IoT. In this paper, we investigate the integration of blockchain technology with IoT. We name such synthesis of blockchain and IoT as blockchain of things (BCoT). This paper presents an in-depth survey of BCoT and discusses the insights of this new paradigm. In particular, we first briefly introduce IoT and discuss the challenges of IoT. Then, we give an overview of blockchain technology. We next concentrate on introducing the convergence of blockchain and IoT and presenting the proposal of BCoT architecture. We further discuss the issues about using blockchain for fifth generation beyond in IoT as well as industrial applications of BCoT. Finally, we outline the open research directions in this promising area.

654 citations


Cites background from "Mobile Edge Computing: A Survey"

  • ...The work of [66] presents an offloading method with consideration of load balancing among multiple MEC servers....

    [...]

  • ...Network slicing can essentially offer a solution to the diverse demands of blockchain applications in MEC....

    [...]

  • ...As a result, MEC can improve the response, privacypreservation and context-awareness in contrast to cloud computing....

    [...]

  • ...However, it is necessary to optimize and allocate both network and computing resources to fulfill the diverse demands in the composite environment of MEC and cloud computing....

    [...]

  • ...For example, cloud servers or some MEC servers may serve as full nodes that store the whole blockchain data and participate in most of blockchain operations, such as initiating transactions, validating transactions (i.e., mining) while IoT devices may serve as lightweight nodes that only store partial blockchain data (even hash value of blockchain data) and undertake some less-computationalintensive tasks (such as initiating transactions) [103]....

    [...]

References
More filters
ReportDOI
28 Sep 2011
TL;DR: This cloud model promotes availability and is composed of five essential characteristics, three service models, and four deployment models.
Abstract: Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model promotes availability and is composed of five essential characteristics, three service models, and four deployment models.

15,145 citations


Additional excerpts

  • ...resources in a mobile environment [21]....

    [...]

Journal ArticleDOI
TL;DR: An overview of the Internet of Things with emphasis on enabling technologies, protocols, and application issues, and some of the key IoT challenges presented in the recent literature are provided and a summary of related research work is provided.
Abstract: This paper provides an overview of the Internet of Things (IoT) with emphasis on enabling technologies, protocols, and application issues. The IoT is enabled by the latest developments in RFID, smart sensors, communication technologies, and Internet protocols. The basic premise is to have smart sensors collaborate directly without human involvement to deliver a new class of applications. The current revolution in Internet, mobile, and machine-to-machine (M2M) technologies can be seen as the first phase of the IoT. In the coming years, the IoT is expected to bridge diverse technologies to enable new applications by connecting physical objects together in support of intelligent decision making. This paper starts by providing a horizontal overview of the IoT. Then, we give an overview of some technical details that pertain to the IoT enabling technologies, protocols, and applications. Compared to other survey papers in the field, our objective is to provide a more thorough summary of the most relevant protocols and application issues to enable researchers and application developers to get up to speed quickly on how the different protocols fit together to deliver desired functionalities without having to go through RFCs and the standards specifications. We also provide an overview of some of the key IoT challenges presented in the recent literature and provide a summary of related research work. Moreover, we explore the relation between the IoT and other emerging technologies including big data analytics and cloud and fog computing. We also present the need for better horizontal integration among IoT services. Finally, we present detailed service use-cases to illustrate how the different protocols presented in the paper fit together to deliver desired IoT services.

6,131 citations


"Mobile Edge Computing: A Survey" refers background in this paper

  • ...useful information that may benefit to different business segments [58]....

    [...]

Journal ArticleDOI
TL;DR: Author(s): Go, Alan S; Mozaffarian, Dariush; Roger, Veronique L; Benjamin, Emelia J; Berry, Jarett D; Blaha, Michael J; Dai, Shifan; Ford, Earl S; Fox, Caroline S; Franco, Sheila; Fullerton, Heather J; Gillespie, Cathleen; Hailpern, Susan M; Heit, John A; Howard, Virginia J; Huffman, Mark D; Judd
Abstract: Author(s): Go, Alan S; Mozaffarian, Dariush; Roger, Veronique L; Benjamin, Emelia J; Berry, Jarett D; Blaha, Michael J; Dai, Shifan; Ford, Earl S; Fox, Caroline S; Franco, Sheila; Fullerton, Heather J; Gillespie, Cathleen; Hailpern, Susan M; Heit, John A; Howard, Virginia J; Huffman, Mark D; Judd, Suzanne E; Kissela, Brett M; Kittner, Steven J; Lackland, Daniel T; Lichtman, Judith H; Lisabeth, Lynda D; Mackey, Rachel H; Magid, David J; Marcus, Gregory M; Marelli, Ariane; Matchar, David B; McGuire, Darren K; Mohler, Emile R; Moy, Claudia S; Mussolino, Michael E; Neumar, Robert W; Nichol, Graham; Pandey, Dilip K; Paynter, Nina P; Reeves, Matthew J; Sorlie, Paul D; Stein, Joel; Towfighi, Amytis; Turan, Tanya N; Virani, Salim S; Wong, Nathan D; Woo, Daniel; Turner, Melanie B; American Heart Association Statistics Committee and Stroke Statistics Subcommittee

4,969 citations

Journal ArticleDOI
TL;DR: This work refers one to the original survey for descriptions of potential applications, summaries of AR system characteristics, and an introduction to the crucial problem of registration, including sources of registration error and error-reduction strategies.
Abstract: In 1997, Azuma published a survey on augmented reality (AR). Our goal is to complement, rather than replace, the original survey by presenting representative examples of the new advances. We refer one to the original survey for descriptions of potential applications (such as medical visualization, maintenance and repair of complex equipment, annotation, and path planning); summaries of AR system characteristics (such as the advantages and disadvantages of optical and video approaches to blending virtual and real, problems in display focus and contrast, and system portability); and an introduction to the crucial problem of registration, including sources of registration error and error-reduction strategies.

3,624 citations

Journal ArticleDOI
TL;DR: The results from a proof-of-concept prototype suggest that VM technology can indeed help meet the need for rapid customization of infrastructure for diverse applications, and this article discusses the technical obstacles to these transformations and proposes a new architecture for overcoming them.
Abstract: Mobile computing continuously evolve through the sustained effort of many researchers. It seamlessly augments users' cognitive abilities via compute-intensive capabilities such as speech recognition, natural language processing, etc. By thus empowering mobile users, we could transform many areas of human activity. This article discusses the technical obstacles to these transformations and proposes a new architecture for overcoming them. In this architecture, a mobile user exploits virtual machine (VM) technology to rapidly instantiate customized service software on a nearby cloudlet and then uses that service over a wireless LAN; the mobile device typically functions as a thin client with respect to the service. A cloudlet is a trusted, resource-rich computer or cluster of computers that's well-connected to the Internet and available for use by nearby mobile devices. Our strategy of leveraging transiently customized proximate infrastructure as a mobile device moves with its user through the physical world is called cloudlet-based, resource-rich, mobile computing. Crisp interactive response, which is essential for seamless augmentation of human cognition, is easily achieved in this architecture because of the cloudlet's physical proximity and one-hop network latency. Using a cloudlet also simplifies the challenge of meeting the peak bandwidth demand of multiple users interactively generating and receiving media such as high-definition video and high-resolution images. Rapid customization of infrastructure for diverse applications emerges as a critical requirement, and our results from a proof-of-concept prototype suggest that VM technology can indeed help meet this requirement.

3,599 citations