ETD system

Electronic theses and dissertations repository


Tesi etd-04162014-094357

Thesis type
Tesi di laurea magistrale
Interference-Aware Mode Selection and Resource Allocation for Device-To-Device (D2D) communications in LTE-Advanced Networks
Corso di studi
relatore Dott. Stea, Giovanni
relatore Prof. Mingozzi, Enzo
relatore Ing. Virdis, Antonio
Parole chiave
  • mode selection
  • lte
  • d2d
  • device-to-device
  • lte-advanced
  • resource allocation
Data inizio appello
Riassunto analitico
The recent socio-technological trend in proximity-based applications and services, and the increasing market interest in short-range communications have triggered the research and standardization communities to explore the potential of device-to-device (D2D) communications in LTE-Advanced networks (LTE Release 12).
D2D communications imply that wireless devices in proximity of each other communicate in a peer-to-peer fashion without the aid of the base station.
This new type of communication holds the promise of three types of gains: High bit-rates, low delays and reduced power consumption from both the users (transmitting at lower powers) and the infrastructure (offloading transmissions to users). In addition, if reuse of the same frequency resources by the D2D devices is enabled, lower use of both uplink and downlink resources is expected. New challenges are also introduced (such as interference management) and effective and efficient solutions have to be developed to deal with them.
In this work, we present a comprehensive framework for D2D transmissions. The latter consists of two interrelated building blocks: an Interference-Aware Mode Selection algorithm, which decides whether pairs of UE should communicate via the infrastructure or directly, based on their mutual interference and channel quality in both directions; a Resource Allocation algorithm, that manages the uplink and downlink frames by scheduling mutually non interfering D2D pairs in shared resources, so as to minimize the number of allocated frequencies.
Simulation results show that the proposed solution enhances both system and end-users performance: more specifically, at low loads we achieve the same throughput with a lower resource (hence power) consumption, and, at higher loads, we move further to the right the saturation limit.