Final Presentations: Monday,  May 19, 2008,  9:00 AM - 1:00 PM,   320 Soda Hall

April 7th:    Please talk to Prof. Tse this week during office hours. Project proposals are due on April 14.    Scroll down and click on "Project List" for project details.

Tuesday, April 1:   I would like to make a correction to the rules of the midterm. **The reference materials allowed for the exam are only the textbook and any material on the 224B website. In particular, your class notes are NOTallowed. This is to make it fair for all students.**

Please pick up your exam at 3:30 from Therese on Tues.
David

There are no office hours today, Monday, March 31, Professor Tse is out of town.
Office hours will resume on Wednesday, April 2, at 11:00 AM.

The 224B take home midterm will be available from Therese on Tues at 3:30 pm. It is due at 3:30 pm on Wed with Therese. There is no lecture on Tues. Here are some rules of the game:

The only references you are allowed are: 1) textbook, 2) powerpoint lecture slides on the class webpage, 3) your class notes, 4) homework solutions on the class webpage. No other references are allowed.

There is absolutely no collaboration or discussions with others, in the class or outside the class. Anyone caught will be given 0 marks for the entire exam. This has happened before.

You should review the material before the exam starts rather than during the exam. Since it is a take home exam, the questions are demanding and some may be a bit open ended. It will assume you already have in-depth knowledge of the material and there is no time for you to study it during the 24 hours of the exam.

In Chapter 5, (see below: Lecture Slides/Transparencies) the material up to pg. 18 is for the midterm.

Good luck!
David

Tuesday, February 19:  Due to a conflict with the Netcom seminar today, Prof. Tse’s office hours have changed today (Tuesday) from 1:00-2:00 to 2:00-3:00 PM.
Further, Tuesday’s office hours will from now on, be held on Monday, 4:00-5:00 PM.  Wednesday office hours, 11:00-12:00 PM will remain unchanged.

HW1 and HW2 grading assignments are due in class Tuesday, February 19.  

Office Hours for Tuesday, February 5, have been changed to 2-3 PM

---

Administrative Info

Office hours:   Monday 4:00 - 5:00 PM  and  Wednesday 11:00 AM - 12:00 PM.

Course Assistant:   Therese George, Room 253 Cory Hall, 642-2384 therese@eecs.berkeley.edu

---

Course Info

Description:

• Course Outline

---

Projects & Presentations

Project List

• Title:  Power Scaling Through Decentralized Multiband Cooperation
• Report
• Slides
• Authors:  Ricardo Garcia and Kristen Woyach
• Abstract:  Consider a cognitive radio scenario in which the secondary users are constrained to only talk when the primary user is unaffected. With an accurate estimate of the primary’s signal power, the secondaries can scale their transmit power to take full advantage of any opportunities. However, accurate estimation proves difficult because of hardware limitations and channel uncertainty. In this project, we explore ways to use any information at our disposal to improve our estimation.
  In current literature, observation of multiple bands provides binary decisions about whether the secondary is being shadowed, which is then shared by all nodes to make a collective decision of whether to transmit. We begin by extending this result to allow individual secondary nodes to quantitatively estimate the severity of their shadowing and scale their transmit power accordingly. We then allow the nodes to cooperate in an effort to improve the estimates of both shadowing and fading and to maximize transmit power in the system as a whole. Finally, we explore a realistic scenario in which any cooperation must occur within the band of interest without the use of a dedicated control channel.

• Title:  A Performance and Power Characterization of a 60GHz Transceiver
• Report
• Slides
• Authors:  Ji-Hoon Park and Dusan Stepanovic
• Abstract:  Driven by the need for more bandwidth and the advance of CMOS RF circuits, the development of the 60 GHz communication system is gaining momentum recently. In this project, we focused on a balanced design of the baseband of this system in terms of the BER performance and power consumption. First, we investigated the specification and channel models proposed for this system. Then, we simulated various transceiver structures and characterized their BER performance and power consumption. The structures were compared based on the results and a conclusion was drawn.

• Title:  A New Paradigm of Interference Management for Cellular Networks
• Report
• Slides
• Authors:  Changho Suh and I-Hsiang Wang
• Abstract:  In this term project, we propose a new way of interference management for cellular networks. To handle interference issues, two traditional schemes have been employed so far:
  (1) orthogonalization(especially for cell-edge users);
  (2) treating interference as noise (for cell-interior users). The natural question is "can we do better than these?''.
  To answer this question, we first introduce a new channel model called interfering multiple access channel (IMAC), in which there are multiple groups (cells) and transmitter (mobile) communicates only with intended receiver (desired base station) in its own group. For 2-IMAC (two cells), we show that the degree-of-freedom (dof) is 2K/(K+1) which approaches to interference-free dof as the number K of mobiles in each cell increases. To develop the achievable scheme, we exploit a recently introduced concept, called interference alignment, in which transmit signals are designed to be aligned into small dimensions at the non-intended receiver. As for sum capacity, we find the outer bound and inner bound (proposed interference alignment scheme) for a single-path channel with propagation delay. It turns out that the gap between them is ~ 2K/(K+1) log(K+1) bits/s/Hz for all SNR when the bandwidth is sufficiently large.

• Title:  Power Control in Wireless Networks
• Report
• Slides
• Authors:  Michael Krishnan and Youwei Zhang
• Abstract:  In [1], Huang et.al. show that in an ad-hoc wireless network in which pairs of nodes are constantly streaming traffic, transmit powers can be tuned in a distribued way to optimize some function of SINRs. But application of this to real networks, requires nodes to share information via broadcast, which can be very inefficient. In [2], Abouei et.al. avoid this problem by proposing an algorithm in which nodes have an assumption about the distribution of the adjacent channel gains and optimize the expected total rate with respect to this distribution. This is a naturally a better approach when the channel is fast-fading, but will achieve worse performance than the Huang algorithm when the fading is slower, since it does not use all the information which could potentially be available. It is also sensitive to the assumptions of the channel distribution. In this project, we compare these two approaches by examining the relative costs of their weaknesses and determine in what types of scenarios each is preferable.

  REFERENCES:
  [1] J. Huang, R. Berry, and M. Honig, “Distributed Interference Compensation for Wireless Networks”, IEEE Journal on Selected Areas in Comm., May 2006.
  [2] J. Abouei, A. Bayesteh, M. Ebrahimi, and A. Khandani, “On the Throughput Maximization in Decentralized Wireless Networks”, to be submitted to IEEE Transactions on Info. Theory, Jan. 2008.

• Title:  Cooperative Spatial Multiplexing in Wireless Networks: A Proposed Scheme Using Limited Feedback
• Report
• Slides
• Author:  Vinayak Nagpal
• Abstract:  In this paper I present a scheme that has potential to exploit degrees of freedom provided by relays for spatial multiplexing. The scheme which relies on limited feedback from destination does not require explicit node discovery and adaptively scales for diversity-spatial multiplexing gains depending on channel realization. I present the motivation for the scheme, its problems and possible directions to address these problems.

• Title:  Structured Codes for Uplink Multiple Cell-Site Processing with Limited Backhaul Capacity
• Report
• Slides
• Author:  Bobak Nazer and Galen Reeves
• Abstract:  In this report, we examine uplink communication in cellular networks. We use a system abstraction proposed by Wyner which places cell-sites on a circle and assumes that each cell-site can only see signals emanating from at most one cell-site away. We are interested in the rates uplink rates achievable through joint decoding of cell-site observations at a central processor. The capacity is known if the observations are available ``for free'' at the central processor. More realistically, the links to the central processor will have their own communication constraints. Here, we improve upon the achievable rates derived by Sanderovich \textit{et al.} for this setting through the use of a new, lattice-based communication strategy: compute-and-forward.

• Title:  Efficient Protocol Design for Single Relay Wireless Networks Using Deterministic Model
• Report
• Slides
• Author:  Sameer Pawar and Nima Noorshams
• Abstract:  In this project we study a flat, slow fading single relay channel,with half duplex constraint on each node using deterministic channelmodel. We analyze Dynamic Decode and Forward (DDF) protocol and justify its sub-optimal performance on diversity multiplexing tradeoff (DMT) frontier compared to MISO system. We also propose a novel Dynamic Partial Decode and Forward (PDDF) protocol to improve on the limitations of DDF and conjecture that it achieves optimal i.e., MISO DMT performance. We also make few comments on how these observations can be utilized to generalize this PDDF protocol to N relay case.

• Title:  ALOHA Design for Large Wireless Network
• Report
• Slides
• Author:  Se Yong Park and Mathias Humbert
• Abstract:  This report is constituted of 5 sections. In section 2, we formulate the problems and articulate the assumptions that we made for analytical ease. Section 3 addresses the single-hop direct communication. From its analysis, we derive a numerical boundary between sparse and dense networks. Section 4 considers the relay communication. Section 5 considers cooperative communication. In section 6, we summarize the results and make a conclusion. Outlines of the proofs are given at the end of this report.

• Title:  SNR Walls for Signal Detection in Packet Based Cognitive Radio Network
• Report
• Slides
• Author:  Maryam Vareth and Mehdi Malboubi
• Abstract:  The performance of a detector in any practical system that operates in low SNR environment is limited by the modeling uncertainties. The impact of these uncertainties is quantified by the position of SNR wall below which a detector will fail to be robust, no matter how long it observes the channel. In this paper, the concept of SNR wall is extended to the packet based network by analyzing the performance of two simple detectors namely averaging and radiometer in presence of packet based primary users . We want to show what aspect of our primary user communication model lead to SNR walls for different levels of knowledge of the traffic patterns of the signal to be detected. The result have proposition for wireless spectrum regulators. The context is opportunistically sharing the spectrum with packet based primary users that must be detected in order to avoid casing harmful interference on a channel. It is shown that averaging the signal is a robust test statistic for the detecting the packet based primary user in the presence of noise uncertainty. In other words, there is no SNR wall in the averaging detector due to noise uncertainty when there is no fading. Furthermore, it is shown that the performance of the radiometer (energy detector) not only is limited by the noise uncertainty but also is limited by the the parameters of the traffic model mainly duty cycle. We have shown that it is possible the number of samples required to detect the presence/absence of a packet based primary user goes to infinity before approaching the SNR wall that is the direct result of noise uncertainty and is presented in [5].

---

Lecture Slides/Transparencies

• Chapter 1 and 2     
• Chapter 3     
• Chapter 4     
• Chapter 5     
• Chapter 6     
• Chapter 7     
• Chapters 8 & 10     

---

Homework and Solutions

Homework #1        

• Homework #1 Solutions

Homework #2 / Due Thursday, Feb. 7, 2008 :  Ex. 2.4, Ex. 2.9, Ex. 2.10 in Chapter 2 of Tse & Viswanath   /   Ex. A. 2, Ex. A. 3 in Appendix A of T & V./     

• Homework #2 Solutions

  Homework #3 / Due Thursday, Feb. 14, 2008 :  Ex. 2.15, 16, 17 in Chapter 2.   /   Ex. A. 6 and A. 7 in Appendix A.

• Homework #3 Solutions

  Homework #4 / Due Thurs. Feb 21.   Problems 3.2, 3.4 and 3.5 in T&V.

• Homework #4 Solutions

  Homework #5 / Due Thurs. Feb 28.   Ex. 3.8, parts 1 -3.   Ex. 3.10, 3.11 3.15 parts 1-2, 3.19.   All from T&V.

• Homework #5 Solutions

  Homework #6 / Due Thurs. March 13.   Ex. 3.21, 3.24, 3.27, 3.28 in T&V.

• Homework #6 Solutions

  Homework #7 / Due Fri. March 21. before 4pm with Therese   Ex. 4.10, 4.11, 4.13, 4.14.

• Homework #7 Solutions

  Homework #8 / Due Thurs, May 8 in class  /  Ex. 5.1, 5.13, 5.20, 5.23  /  Ex. 7.2, 7.5

• Homework #8 Solutions