PhD seminar - Hamidreza Ebrahimzadeh SaffarExport this event to calendar

Tuesday, August 12, 2014 — 10:30 AM EDT

Candidate

Hamidreza Ebrahimzadeh Saffar

Title

Asynchronous Joint Source-Channel Communication: An information Theoretic Perspective

Supervisor

Patrick Mitran

Abstract

Due to the increasing growth and demand for wireless communication services, new techniques and paradigms are required for the development of next generation systems and networks. As a first step to better differentiate between various options to develop future systems, one should consider fundamental theoretical problems and limitations in present systems and networks. Hence, some common ground between network information theory and mobile/wireless medium techniques should be explicitly addressed to better understand future generation trends.

Among practical limitations, a major challenge, which is inherent and due to the physics of many mobile/wireless setups, is the problem of asynchronism between different nodes and/or clients in a wireless network. Although analytically convenient, the assumption of full synchronization between the end terminals in a network is usually difficult to justify. Thus, finding fundamental limits for communication systems under different types of asynchronism is essential to tackle real world problems.

In this thesis, we study information theoretic limits that various multiuser wireless communication systems encounter under time or phase asynchronism between different nodes. In particular, we divide our research into two categories: phase asynchronous and time asynchronous systems.

In the first part of this thesis, we consider several multiuser networks with phase fading communication links, i.e., all of the channels introduce phase shifts to the transmitted signals. We assume that the phase shifts are unknown to the transmitters as a practical assumption which results in a phase asynchronism between transmitter sides and receiver sides. We refer to these communication systems as phase incoherent (PI) communication systems and study the problem of communicating arbitrarily correlated sources over them. Specifically, we are interested in solving the general problem of joint source-channel coding over PI networks. To this end, we first present a lemma which is very useful in deriving necessary conditions for reliable communication of the sources over PI channels. Then, for each channel and under specific gain conditions, we derive sufficient conditions based on separate source and channel coding and show that the necessary and sufficient conditions match. Therefore, we are able to present and prove several separation theorems for channels under study under specific gain conditions.

In the second part of this thesis, we consider time asynchronism in networks. In particular, we consider a multiple access channel with a relay as a general setup to model many wireless networks in which the transmitters are time asynchronous in the sense that they cannot operate at the same exact time. Based on the realistic assumption of a time offset between the transmitters, we again consider the problem of communicating arbitrarily correlated sources over such a time-asynchronous multiple access relay channel (TA-MARC). We first derive a general necessary condition for reliable communication. Then, by the use of separate source and channel coding and under specific gain conditions, we show that the derived sufficient conditions match with the general necessary condition for reliable communications. Consequently, we present a separation theorem for this class of networks under specific gain conditions. We then specialize our results to a two-user interference channel with time asynchronism between the encoders

Location 
EIT building
Room 3145

,

S M T W T F S
26
27
28
29
30
31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
1
2
3
4
5
6
  1. 2019 (194)
    1. December (1)
    2. November (1)
    3. October (1)
    4. September (1)
    5. August (26)
    6. July (40)
    7. June (24)
    8. May (23)
    9. April (35)
    10. March (25)
    11. February (9)
    12. January (10)
  2. 2018 (150)
    1. December (13)
    2. November (25)
    3. October (12)
    4. September (13)
    5. August (7)
    6. July (23)
    7. June (9)
    8. May (6)
    9. April (9)
    10. March (16)
    11. February (10)
    12. January (7)
  3. 2017 (212)
  4. 2016 (242)
  5. 2015 (242)
  6. 2014 (268)
  7. 2013 (192)
  8. 2012 (31)