Yingchao Zhao

Task Allocation on Nonvolatile-Memory-Based Hybrid Main Memory

In this paper, we consider the task allocation problem on a hybrid main memory composed of nonvolatile memory (NVM) and dynamic random access memory (DRAM). Compared to the conventional memory technology DRAM, the emerging NVM has excellent energy performance since it consumes orders of magnitude less leakage power. On the other hand, most types of NVMs come with the disadvantages of much shorter write endurance and longer write latency as opposed to DRAM. By leveraging the energy efficiency of NVM and long write endurance of DRAM, this paper explores task allocation techniques on hybrid memory for multiple objectives such as minimizing the energy consumption, extending the lifetime, and minimizing the memory size. The contributions of this paper are twofold. First, we design the integer linear programming (ILP) formulations that can solve different objectives optimally. Then, we propose two sets of heuristic algorithms including three polynomial time offline heuristics and three online heuristics. Experiments show that compared to the optimal solutions generated by the ILP formulations, the offline heuristics can produce near-optimal results.

Barrier Coverage by Sensors with Adjustable Ranges

One of the most fundamental tasks of wireless sensor networks is to provide coverage of the deployment region. We study the coverage of a line interval with a set of wireless sensors with adjustable coverage ranges. Each coverage range of a sensor is an interval centered at that sensor whose length is decided by the power the sensor chooses. The objective is to find a range assignment with the minimum cost. There are two variants of the optimization problem. In the discrete variant, each sensor can only choose from a finite set of powers, whereas in the continuous variant, each sensor can choose power from a given interval. For the discrete variant of the problem, a polynomial-time exact algorithm is designed. For the continuous variant of the problem, NP-hardness of the problem is proved and followed by an ILP formulation. Then, constant-approximation algorithms are designed when the cost for all sensors is proportional to rκ for some constant κ ≥ 1, where r is the covering radius corresponding to the chosen power. Specifically, if κ = 1, we give a 1.25-approximation algorithm and a fully polynomial-time approximation scheme; if κ > 1, we give a 2-approximation algorithm. We also show that the approximation analyses are tight.

Power-Aware Variable Partitioning for DSPs With Hybrid PRAM and DRAM Main Memory

Phase change random access memory (PRAM) is one kind of nonvolatile memory, which is desirable to be used for DSP systems as main memory, as it consumes less power than DRAM and is much denser than DRAM. In this paper, we utilize a hybrid main memory composed of DRAM and PRAM, which leverages the low power consumption of PRAM while minimizing the performance and lifetime degradation caused by PRAM write. To make full use of different advantages of DRAM and PRAM, especially for the application-specific DSP systems, we reconsider the variable partitioning and instruction scheduling problems on the hybrid main memory. Different optimization objectives, for example power consumption, schedule length, and the number of writes on PRAM, are considered. At the same time, different kinds of hybrid architectures are analyzed. Graph models, ILP model, and algorithms are proposed for different settings. Experiments show that the proposed techniques reduce up to 49% power consumption and 88% the number of writes on PRAM on average.

The Isolation Game: A Game of Distances

We introduce a new multi-player geometric game, which we will refer to as the isolation game, and study its Nash equilibria and best or better response dynamics. The isolation game is inspired by the Voronoi game, competitive facility location, and geometric sampling. In the Voronoi game studied by Durr and Thang, each players objective is to maximize the area of her Voronoi region. In contrast, in the isolation game, each players objective is to position herself as far away from other players as possible in a bounded space. Even though this game has a simple definition, we show that its game-theoretic behaviors are quite rich and complex. We consider various measures of farness from one player to a group of players and analyze their impacts to the existence of Nash equilibria and to the convergence of the best or better response dynamics: We prove that it is NP-hard to decide whether a Nash equilibrium exists, using either a very simple farness measure in an asymmetric space or a slightly more sophisticated farness measure in a symmetric space. Complementing to these hardness results, we establish existence theorems for several special families of farness measures in symmetric spaces: We prove that for isolation games where each player wants to maximize her distance to her m th nearest neighbor, for any m, equilibria always exist. Moreover, there is always a better response sequence starting from any configuration that leads to a Nash equilibrium. We show that when m = 1 the game is a potential game -- no better response sequence has a cycle, but when m > 1 the games are not potential. More generally, we study farness functions that give different weights to a players distances to others based on the distance rankings, and obtain both existence and hardness results when the weights are monotonically increasing or decreasing. Finally, we present results on the hardness of computing best responses when the space has a compact representation as a hypercube.

Maximizing common idle time on multi-core processors with shared memory

Reducing energy consumption is a critical problem in most of the computing systems today. This paper focuses on reducing the energy consumption of the shared main memory in multi-core processors by putting it into sleep state when all the cores are idle. Based on this idea, this work presents systematic analysis of different assignment and scheduling models and proposes a series of scheduling schemes to maximize the common idle time of all cores. An optimal scheduling scheme is proposed assuming the number of cores is unbounded. When the number of cores is bounded, an efficient heuristic algorithm is proposed. The experimental results show that the heuristic algorithm works efficiently and can save as much as 25.6% memory energy compared to a conventional multi-core scheduling scheme.

Energy-aware register file re-partitioning for clustered VLIW architectures

VLIW architectures have gained acceptance in embedded systems. Traditional monolithic register file is not suitable for VLIW architectures with a large number of functional units. Clustered VLIW architecture is often applied, where the register file is partitioned into a number of smaller register files. Register files represent a substantial portion of the energy consumption in modern processors, and it is growing rapidly with wider instruction width. Most of the known clustered VLIW architectures partition the register file evenly among clusters. In this paper, we study the effect of energy consumption with register file re-partitioning on clustered VLIW architecture, where register files are not necessarily partitioned evenly. We present algorithms to compute energy-efficient re-partition of register files under different conditions. The impact of different intercluster communication models as well as the impact of program behavior on the register file re-partitioning are analyzed in this paper. Experimental results show that energy saving can be achieved using the proposed techniques.

The isolation game: A game of distances

We introduce a new multi-player geometric game, which we will refer to as the isolation game, and study its Nash equilibria and best or better response dynamics. The isolation game is inspired by the Voronoi game, competitive facility location, and geometric sampling. In the Voronoi game studied by Durr and Thang, each players objective is to maximize the area of her Voronoi region. In contrast, in the isolation game, each players objective is to position herself as far away from other players as possible in a bounded space. Even though this game has a simple definition, we show that its game-theoretic behaviors are quite rich and complex. We consider various measures of farness from one player to a group of players and analyze their impacts to the existence of Nash equilibria and to the convergence of the best or better response dynamics: We prove that it is NP-hard to decide whether a Nash equilibrium exists, using either a very simple farness measure in an asymmetric space or a slightly more sophisticated farness measure in a symmetric space. Complementing these hardness results, we establish existence theorems for several special families of farness measures in symmetric spaces: We prove that, for isolation games where each player wants to maximize her distance to her mth nearest neighbor, for any m, equilibria always exist. Moreover, there is always a better response sequence starting from any configuration that leads to a Nash equilibrium. We show that when m=1 the game is a potential game, no better response sequence has a cycle, but when m>1 the games are not potential. More generally, we study farness functions that give different weights to a players distances to others based on the distance rankings, and obtain both existence and hardness results when the weights are monotonically increasing or decreasing. Finally, we present results on the hardness of computing best responses when the space has a compact representation as a hypercube.

Combinatorial and spectral aspects of nearest neighbor graphs in doubling dimensional and nearly-Euclidean spaces

Miller, Teng, Thurston, and Vavasis proved that every k- nearest neighbor graph (k-NNG) in Rdhas a balanced vertex separator of size O(n1-1/dk1/d). Later, Spielman and Teng proved that the Fiedler value -- the second smallest eigenvalue of the graph -- of the Laplacian matrix of a k-NNG in Rd is at O(1/n2/d). In this paper, we extend these two results to nearest neighbor graphs in a metric space with doubling dimension γ and in nearly-Euclidean spaces. We prove that for every l > 0, each k-NNG in a metric space with doubling dimension γ has a vertex separator of size O(k2l(32l + 8)2γ log2 L/S log n + n/l), where L and S are respectively the maximum and minimum distances between any two points in P, and P is the point set that constitutes the metric space. We show how to use the singular value decomposition method to approximate a k-NNG in a nearly-Euclidean space by an Euclidean k-NNG. This approximation enables us to obtain an upper bound on the Fiedler value of the k-NNG in a nearly-Euclidean space.

On Approximation Ratios of Minimum-Energy Multicast Routing in Wireless Networks

In the broadcasting of ad hoc wireless networks, energy conservation is a critical issue. Three heuristic algorithms were proposed in Wieselthier et al.(2001) for finding approximate minimum-energy broadcast routings: MST(minimum spanning tree), SPT(shortest-path tree), and BIP(broadcasting incremental power). Wan et al.(2001) characterized their performance in terms of approximation ratios. This paper points out some mistakes in the result of Wan et al.(2001), and proves that the upper bound of sum of squares of lengths of the edges in Euclidean MST in unit disk can be improved to 10.86, thus improves the approximation ratios of MST and BIP algorithm.

Barrier Coverage Using Sensors with Offsets

One of the most fundamental tasks of wireless sensor networks is to provide coverage of barrier, which focuses on detecting intruders crossing a specific region. Suppose that all sensors are dropped from an aircraft along a given line interval, and each sensor has circular coverage range of arbitrary radii. Due to the environmental factors, the sensors will be distributed along the deployment line interval with random offsets. We study the barrier coverage problem with line-based offsets deployments by a set of wireless sensors with adjustable coverage ranges. The objective is to find a range assignment with the minimum cost. In this paper, we present a constant-approximation algorithm and two fully polynomial time approximation schemes (FPTASes) for the barrier coverage by using sensors with offsets under a linear cost function on the sensors range. We also show the performance of the approximation algorithms by experiments.