____________________________________________________________________

Title: Exact Byzantine Consensus in Directed Graphs
Abstract: 

For synchronous point-to-point n-node networks of undirected links, it has been previously shown that, to achieve consensus in presence of up to f Byzantine faults, the following two conditions are together necessary and sufficient: (i) n >= 3f + 1 and (ii) network connectivity greater than 2f. The first condition, that is, n >= 3f + 1, is known to be necessary for directed graphs as well. On the other hand, the second condition on connectivity is not necessary for directed graphs. So far, tight necessary and sufficient condition for Byzantine consensus in directed graphs has not been developed. This paper presents tight necessary and sufficient condition for achieving Byzantine consensus in synchronous networks that can be represented as directed graphs. We provide a constructive proof of sufficiency by presenting a new Byzantine consensus algorithm for directed graphs. Further work is needed to improve the message overhead of Byzantine consensus in directed graphs.

____________________________________________________________________

Title: Beyond-CMOS Circuit Design and Spin-Diode Logic

Abstract:

A variety of newly available materials and devices have been evaluated as building blocks for next generation computing. These devices can be classified into three broad groups: non-binary logic families (e.g., quantum computing), binary logic families with a CMOS logic structure (e.g., graphene FETs and carbon nanotube FETs), and binary logic families with a novel logic structure (e.g., spintronics, molecular electronics, and nanowires). There are significant challenges preventing the implementation of all of these logic families, and none of these technologies are currently ready for use on a large scale. Cascading, one device directly driving another device, is a particularly significant challenge, and few emerging technologies have demonstrated cascaded circuits. It is therefore interesting to consider spin-diode logic as an example, a logic family for which cascading is a primary design consideration.

____________________________________________________________________

Title: Distributed Optimal Power Flow for Smart Microgrids
Abstract:

Optimal power flow (OPF) is considered for microgrids, with the objective of minimizing
either the power distribution losses, or, the cost of power drawn from the substation and
supplied by distributed generation (DG) units. The microgrid is unbalanced, due to unequal loads in each phase and non-equilateral conductor spacings on the distribution lines. Similar to OPF formulations for balanced systems, the considered OPF problem is nonconvex. A semidefinite relaxation (SDR) technique is advocated to obtain a convex problem with improved performance. To ensure scalability, robustness, and data privacy, the relaxed problem is further solved in a distributed fashion by resorting to the alternating direction method of multipliers. The resulting algorithm entails iterative message-passing among groups of consumers with faster convergence rate compared to competing alternatives.

____________________________________________________________________

Title: Distributed Estimation
Abstract: 

Multi-agent systems will be used for distributed estimation tasks in the near future. We first survey the distributed estimation and consensus algorithms currently discussed in literature. Prior consensus algorithms do not incorporate nonlinear target dynamic models or higher order moments of the estimate of the target. These limitations are addressed using the Bayesian consensus filters (BCF). Moreover, the BCF using the logarithmic opinion pool yields the optimum global estimate; hence it becomes a distributed estimator. Finally, computational methods for implementing BCF are discussed.

 ____________________________________________________________________

Title: A Predictive Framework for Duty-Cycle Adaptation Using Object Tracking for Wireless Camera Networks 

Abstract:

Energy efficiency dominates practically every aspect of the design of a wireless sensor
network and duty cycling is an important tool for achieving high energy efficiencies. Duty cycling for wireless camera networks meant for tracking objects is made complex by the nodes having to anticipate the arrival of the objects in their field of view. The consequences of an object arriving in the view-region of a camera when it is sleeping need no elaboration. Our work presents a predictive framework to provide nodes with an ability to anticipate the arrival of objects in the field-of-view of their cameras. Our predictive framework differs from others in that the nodes whose duty cycles are increased are at least one step removed from the immediate neighborhood of the nodes where the objects are currently visible. By eliminating the need for the currently busiest nodes to also be in charge of informing their non-busy immediate neighbors to get ready for object arrival, we end up with a more robust strategy for updating the duty-cycle at the nodes where the objects are highly likely to appear soon. The proposed scheme does not require any modification to the structures of the existing MAC protocols. Our contribution includes evaluations based on large-scale simulations and with real data in a iMote2-based testbed.

____________________________________________________________________

Title:  ShareRI.org: a social network for the research community to make research more interactive and fun! 

Abstract:

ShareRI.org, short for Share Research Ideas, is a social network I built in summer 2012 with the aim of making research more interactive and fun. In this nontechnical talk, I will share with you some interesting stories on how I built this website, and why it is useful for the research community, especially for PhD students. For more information, welcome to visit the website: http://shareri.org Bio: Quan Geng is a fourth year graduate student in department of ECE, UIUC. He received Bachelor degree in EE from Tsinghua University in 2009, and M.S. in ECE and M.S. in Mathematics from UIUC in 2011 and 2012. He is currently pursuing M.S. in CS and Ph.D. in ECE, and working with Prof. Pramod Viswanath on wireless communication, information theory and differential privacy in CSL. Since June 2012, he has been serving as the President of Tsinghua University Alumni Association in UIUC.

____________________________________________________________________

Title: Reachability Analysis of Nonlinear Analog Circuits through Iterative Reachable Set Reduction

Abstract:

We propose a methodology for reachability analysis of nonlinear analog circuits to verify
safety properties. Our iterative reachable set reduction algorithm initially considers the entire state space as reachable. Our algorithm iteratively determines which regions in the state space are unreachable and removes those unreachable regions from the over approximated reachable set. We use the State Partitioning Tree (SPT) algorithm to recursively partition the reachable set into convex polytopes. We determine the reachability of adjacent neighbor polytopes by analyzing the direction of state space trajectories at the common faces between two adjacent polytopes. We model the direction of the trajectories as a reachability decision function that we solve using a sound root counting method. We are faithful to the nonlinearities of the system. We demonstrate the memory efficiency of our algorithm through computation of the reachable set of Van der Pol oscillation circuit.

___________________________________________________________________

Title : Dynamic Interference Management

Abstract:

We consider a linear interference network with K transmitter-receiver pairs, where each transmitter can be connected to two neighboring receivers. Long-term fluctuations (shadow fading) in the wireless channel can lead to any link being erased with probability p. Each receiver is interested in one unique message that can be available at M transmitters. In a cellular downlink scenario, the case where M=1 reflects the cell association problem, and the case where M>1 reflects the problem of setting up the backhaul links for Coordinated Multi-Point (CoMP) transmission. In both cases, we analyze Degrees of Freedom (DoF) optimal schemes for the case of no erasures, and propose new schemes with better average DoF performance at high probabilities of erasure. The proposed schemes suggest that the knowledge of the probability of erasure p is necessary for the selection of a message assignment that maximizes the average DoF. Also, a new role of cooperation is identified in dynamic interference networks. The availability of a message at more than one transmitter may not only be used to cancel its interference at other receivers, but to increase the chances of delivering it to its designated receiver.

____________________________________________________________________

Enrique Carrion
Title: Pulsed Nanosecond Characterization of Graphene Transistors
Abstract:

We study hysteresis of top-gated graphene field effect transistors (GFETs) with nanosecond range pulsed gate and drain voltages. The drain current degrades over short time scales, suggesting the presence of at least two charge trapping mechanisms with different time constants (~1 and 10 s). We demonstrate hysteresis-free transfer characteristics of a large number of devices with pulses as short as 100 ns at the drain and gate terminals. The pulsed, hysteresis-free operation enables reliable extraction of graphene mobilities which are independent of sweep direction and are up to a factor of two higher than those obtained from DC characterization. We also observe drain induced trapping effects when lateral fields applied are higher than ~0.3 V/m. Our study illustrates important time and field dependencies necessary to take into account when operating top-gated GFETs in ambient conditions.

____________________________________________________________________

Nisha Somnath
Title: On Deciding the Existence of a Liveness Enforcing Supervisory Policy in a Class of Partially- Controlled General Free-Choice Petri Nets.
Abstract:

If there are transitions in a PN that cannot be prevented from firing by a supervisory
policy, then we have a partially-controlled PN. The existence of a liveness enforcing
supervisory policy (LESP) in a partially-controlled ordinary PN is undecidable. Consequently, there can be no algorithms that synthesize an LESP for an arbitrary ordinary (general) PN. In contrast, we identify a class of general Free-Choice PN (FCPN) structures, which strictly includes the class of ordinary FCPN structures, where the existence of an LESP in any marked member of the class is decidable. Note to Practitioners Every computer user has encountered situations where an unresponsive program enters into a state of suspended animation for perpetuity. A scheduling policy that can guarantee livelock-freedom is highly desirable in this, and other instances of concurrent systems. The results of this paper show that if the concurrent system is modeled as a Petri net that belongs to the class identified in this paper, then it is possible to determine if there is a policy that avoids livelocks, which paves the way for algorithm-development for livelock-avoidance.

 ____________________________________________________________

Leila Fuladi
Title: A tutorial presentation on Polar Codes.
Abstract:

Polar codes, recently proposed by Erdal Arikan, are of theoretical importance because they provably achieve channel capacity with low-complexity encoding and decoding. In this presentation, the speaker will cover the basic theory and will review some theorems on channel polarization and polar coding.

____________________________________________________________________

Qiaomin Xie
Title:  Join-Idle-Queue: A novel load balancing algorithm for dynamically scalable web services
Abstract:

An increasing variety of dynamic-content web services, such as search, social networking and on-line commerce, have been moved to the Cloud service data centers. Two distinct challenges of the Cloud, scalability and large-scale front end, make the design of distributed load balancers highly desirable. Existing algorithms with centralized design, such as Join-the-Shortest-Queue (JSQ), while achieving near-optimality, introduce overwhelming communication between distributed dispatchers and servers. However, approaches based on randomized algorithm, which requires little communication, have performance far from optimality. We proposed a novel class of load balancing algorithms called Join-Idle-Queue (JIQ) for large-scale system with distributed dispatchers. Our previous analysis and evaluation of the JIQ algorithm showed that it significantly outperforms existing algorithms in homogeneous systems. Moreover, the JIQ algorithms require little extra communication on the critical path of request assignment.

_____________________________________________________________

 James Yang
Title : Single Video Performance Analysis for Video-on-Demand Systems
Abstract:

Video-on-Demand system in which we primarily study the content placement problem in
cache delivery systems. Given a closed system with fixed network graph and sufficient
bandwidth, the storage constraint dictates which placement policy is optimal. We narrow
our focus on specific videos of different popularity to analyze, compare and provide bounds on adaptive/non-adaptive and fractional/integer video storage placement policies. Then, we comment on the efficiencies of the optimal placement policy for the whole system and provide an alternate mixed strategy that is close to optimal and efficient.

____________________________________________________________________

Long Le
Title: Energy-efficient detection system in time-varying signal and noise power
Abstract:

In many detection applications with battery powered or energy-harvesting sensors,
energy constraints preclude the use of the optimal detector all the time. Optimal energy performance trade-off is therefore needed in such situations. In many applications, the
signal and noise power may vary greatly over time, and this can be exploited to constrain energy consumption while maintaining the best possible performance. A detector scheduling algorithm based on the signal and noise power information is developed in this work. The resulting algorithm is simple due to its threshold-test structure and can be easily implemented with almost no overhead. A detection system with two detectors using the proposed scheduling scheme is estimated to greatly reduce the energy consumption for a wildlife monitoring application.

____________________________________________________________________

Ali Khanafer
Title: Optimal strategies for controlling information spread on networks
Abstract: