Analytical study of a service discovery system based on an LTE-A D2D implementation

Abstract

This work adapts the paradigm of Cloud Computing over the framework of Device-to-Device (D2D) proximity-based communications in order to offer proximity services to LTE network subscribers, and offload major traffic from the core network. Our proposed work introduces a platform in which mobile devices can offer network services to other nearby devices, and thus act as service end points, or mobile cloudlets. Devices seeking particular services can discover these providers and communicate with them directly, but with the help of the LTE network whose role is limited to assisting in the service provider discovery process. D2D can introduce several positive aspects within any community, through (1) helping tourism by making it seamless to obtain needed services; (2) improving social ties among the society members; and (3) helping the economy through creating a more conducive environment for thriving personal-level businesses. From an implementation point of view, our solution is in line with the 3GPP standards, utilizes the 3GPP-proposed network elements designated for offering proximity based services, and introduces no changes to the rest of the LTE system. Our experimental results proved the effectiveness of our system in helping mobile users discover needed services offered by nearby providers, and therefore communicate directly in a peer-to-peer fashion.

Introduction

The usage of mobile apps on smart phones and tablets has exploded during the recent years, and will continue to increase exponentially according to the 2015 Mobile Economy Report  [1], which envisions an additional one billion subscribers by 2020. This growth will lead to huge mobile data traffic on the network, thus, raising the flag to find ways to increase the network capacity and accommodate these bandwidth consuming applications and services. The Third-Generation Partnership Project (3GPP) has defined data offloading as a solution to cope with this problem and produced a new Rel-12 item, named “Study on Proximity-Based Services (ProSe)”, to identify the potential requirements for an operator to integrate Device-to-Device (D2D) communication in their networks  [2]. This technology has been proposed as a promising concept to improve user experiences and resource utilization in cellular networks by taking advantage of users’ proximity. That is, with D2D, two mobile devices in proximity of each other, can establish a direct local link and bypass the base station (eNodeB), thus (1) enabling very high bit rates, low delays, and low power consumption  [3]; (2) improving spectrum reuse and system throughput since the radio resources may be simultaneously used by cellular and D2D links; and (3) offering a hop gain since the link in the D2D mode is single rather than using both an uplink and downlink resource  [4]. Researchers working on D2D communication have proposed solutions to challenges, like resource allocation and resource management, to coordinate between cellular and D2D users  [5], [6]. However, only few works were done to propose enhancements at the architecture level for incorporating D2D communication in LTE-A. Accordingly, this paper aims to fill this gap, and makes the following contributions:

• It offers a solution for service registration and peer discovery, as they represent the key functions of a D2D communication framework. Our work fits the LTE architecture and conforms to the relevant standards. We use the same architectural elements, but add few functions and signaling messages to these elements.

• It extends the concept of network services in the LTE-A literature and the use cases in the ProSe standards by integrating cloud computing-like services, and adding the concept of “Provider UE” and “Requester UE”.

• It removes the assumption used in the ProSe standards, whereby a device knows the ID of the other device searching for. In our framework, the UE can search for a desired service, not a specific UE, with the use of keywords, leaving it up to the LTE network to locate devices in proximity that offer the requested services.

• It offers a system that allows providers of arbitrary services (and are subscribers to the cellular system) to advertise their services through the LTE-A system, and other users (also subscribers) to discover such services and consume them through integrating the concept of cloudlets into the D2D framework.

We should mention that our work is restricted to the application and networking levels of service registration and discovery for D2D communication, and does not address the communication aspects such as resource allocation, interference management, communication energy consumption, and network capacity. These communication issues fall outside the scope of this paper, and are the subject of a following work that we are starting to work on.

The rest of this paper is organized as follows: Section  2 covers the related work in the literature and the standards, while Section  3 describes the details of our proposed scheme for service registration and discovery. Section  4 analyzes the system performance, and Section  5 concludes the paper and proposes future related works.