Traffic-Adaptive and Energy-Efficient Small Cell Networks-Energy, Delay and Throughput

Mirza Nazrul Alam

Research output: ThesisDoctoral ThesisMonograph

Abstract

The low power small cell network has emerged as a promising and feasible solution to address the massive wireless traffic resulting from the aggressive growth of wireless applications. It is also estimated that Internet of things (IoT) will consist of around 50 billion physical objects by 2020. As a result, besides capacity enhancement, other challenges, e.g., energy efficiency, dynamic addressing of UL/DL traffic asymmetry, low latency, multi-hop communications, reliability and coverage have become the crucial issues in wireless communication technology. Also, in LTE-A, the introduction of Local IP Address (LIPA), Selected IP Traffic Offload (SIPTO) and IP Flow Mobility (IFOM) have opened the opportunity to integrate WiFi or the non-3GPP devices with femto networks. In the above context, two dominant candidates for small cell technology, i.e., the WiFi (IEEE 802.11s) and LTE TDD femto networks, are studied in terms of energy, delay and throughput in this thesis. Basically, small cells are low power, short range and low cost small base stations (BSs) or access points (APs) that operate in either the licensed or unlicensed spectrum. In this thesis the performance of such small cells is studied both analytically and numerically. The experiments are carried out for both UDP and TCP traffic. For the IEEE 802.11s system, the peer-specific queue and the batch scheduling process for the energy saving MAC are introduced in this thesis. The study suggests that at the cost of delay-throughput, the IEEE 802.11s network can save up to 80% energy. In this multi-hop traffic adaptive system, the delay, throughput and fairness can be adjusted by changing the link specific power save modes as well as the beacon interval. For the licensed spectrum, the scope and feasibility of the dynamic LTE TDD network are examined in detail from the implementation perspective. A flexible frame selection scheme is introduced in this respect. At small load, a micro-sleep based operational framework at symbol level for LTE TDD is proposed in this thesis too. The study reveals that a BS can operate in micro-sleep mode while providing the necessary QoS and during off peak hours, the power amplifier (PA) can save up to 90% energy. Two centralized algorithms and one distributed algorithm, based on HNN, are introduced in the thesis to address the traffic asymmetry dynamically. It is found that the algorithms can add 13 to 20 percent additional capacity. Also the algorithms are portable to the 3GPP LTE TDD system. Regardless of the number of links the algorithms converge within the first few epochs.
Translated title of the contributionTraffic-Adaptive and Energy-Efficient Small Cell Networks-Energy, Delay and Throughput
Original languageEnglish
QualificationDoctor's degree
Awarding Institution
  • Aalto University
Supervisors/Advisors
  • Jäntti, Riku, Supervising Professor
  • Jäntti, Riku, Thesis Advisor
Publisher
Print ISBNs978-952-60-7216-6
Electronic ISBNs978-952-60-7215-9
Publication statusPublished - 2016
MoE publication typeG4 Doctoral dissertation (monograph)

Keywords

  • Small cell
  • 802.11s
  • femtocell
  • dynamic LTE TDD
  • traffic-adaptive
  • DTX
  • DRX
  • HARQ
  • link specific PSM
  • energy-efficient
  • HNN
  • GA

Fingerprint Dive into the research topics of 'Traffic-Adaptive and Energy-Efficient Small Cell Networks-Energy, Delay and Throughput'. Together they form a unique fingerprint.

Cite this