Yuankun Xue

Prediction and control of bursty cloud workloads: a fractal framework

Cloud Computing is a promising approach to handle the growing needs for computation and storage in an efficient and cost-effective manner. Towards this end, characterizing workloads in the cloud infrastructure (e.g., a data center) is essential for performing cloud optimizations such as resource provisioning and energy minimization. However, there is a huge gap between the characteristics of actual workloads (e.g., they tend to be bursty and exhibit fractal behavior) and existing cloud optimization algorithms, which tend to rely on simplistic assumptions about the workloads. To close this gap, based on fractional calculus concepts, we present a fractal model to account for the complex dynamics of cloud computing workloads (i.e., the number of request arrivals or CPU/memory usage during each time interval). More precisely, we introduce a fractal operator to account for the time-varying fractal properties of the cloud workloads. In addition, we present an efficient (online) parameter estimation algorithm, an accurate forecasting strategy, and a novel fractal-based model predictive control approach for optimizing the CPU utilization, and hence, the overall energy consumption in the system while satisfying networked architecture performance constraints like queue capacities. We demonstrate advantages of our fractal model in forecasting the complex cloud computing dynamics over conventional (non-fractal) models by using real-world cloud (Google) traces. Unlike non-fractal models, which have very poor prediction capabilities under bursty workload conditions, our fractal model can accurately predict bursty request processes, which is crucial for cloud computing workload forecasting. Finally, experimental results demonstrate that the fractal model based optimization outperforms the non-fractal based ones in terms of minimizing the resource utilization by an average of 30%.

User Cooperation Network Coding Approach for NoC Performance Improvement

The astonishing rate of sensing modalities and data generation poses a tremendous impact on computing platforms for providing real-time mining and prediction capabilities. We are capable of monitoring thousands of genes and their interactions, but we lack efficient computing platforms for large-scale (exa-scale) data processing. Towards this end, we propose a novel hierarchical Network-on-Chip (NoC) architecture that exploits user-cooperated network coding (NC) concepts for improving system throughput. Our proposed architecture relies on a light-weighted subnet of cooperation unit routers (CUR) for multicast traffic. Coding network interface (CNI) performs encoding/decoding of NC symbols and shares the data flows among cooperation units(CUs). We endow our proposed NC-based NoC architecture with: (i) a corridor routing algorithm (CRA) for maximizing network throughput and (ii) an adaptive flit dropping (AFD) scheme to mitigate congestion, branch-blocking and deadlock at run-time. The experimental results demonstrate that our proposed platform offers up to 127X multicast throughput improvement over multiple-unicast and XY tree-based multicast under synthetic collective traffic scenario. We have evaluated the proposed platform with different realworld benchmarks under network sizes of 4x4 to 32x32. Simulation results show 21%--91% latency improvement and up to 25X runtime reduction over conventional mesh NoC performing genetic-algorithm based protein folding analysis. FPGA implementation results show minimal overhead.

An efficient Network-on-Chip (NoC) based multicore platform for hierarchical parallel genetic algorithms

In this work, we propose a new Network-on-Chip (NoC) architecture for implementing the hierarchical parallel genetic algorithm (HPGA) on a multi-core System-on-Chip (SoC) platform. We first derive the speedup metric of an NoC architecture which directly maps the HPGA onto NoC in order to identify the main sources of performance bottlenecks. Specifically, it is observed that the speedup is mostly affected by the fixed bandwidth that a master processor can use and the low utilization of slave processor cores. Motivated by the theoretical analysis, we propose a new architecture with two multiplexing schemes, namely dynamic injection bandwidth multiplexing (DIBM) and time-division based island multiplexing (TDIM), to improve the speedup and reduce the hardware requirements. Moreover, a task-aware adaptive routing algorithm is designed for the proposed architecture, which can take advantage of the proposed multiplexing schemes to further reduce the hardware overhead. We demonstrate the benefits of our approach using the problem of protein folding prediction, which is a process of importance in biology. Our experimental results show that the proposed NoC architecture achieves up to 240X speedup compared to a single island design. The hardware cost is also reduced by 50% compared to a direct NoC-based HPGA implementation.

A spatio-temporal fractal model for a CPS approach to brain-machine-body interfaces

Capturing the mathematical features of physical and cyber processes is essential for endowing the CPS with built-in intelligence. In this paper, we develop a compact yet accurate mathematical model able to capture the spatio-temporal fractal cross-dependencies between coupled processes and illustrate its benefits within the context of brain-machine-body interface. Our generalized mathematical model improves the modeling accuracy of the dynamics of biological processes and is validated against medical observations.

Multicast-Aware High-Performance Wireless Network-on-Chip Architectures

Today’s multiprocessor platforms employ the network-on-chip (NoC) architecture as the preferable communication backbone. Conventional NoCs are designed predominantly for unicast data exchanges. In such NoCs, the multicast traffic is generally handled by converting each multicast message to multiple unicast transmissions. Hence, applications dominated by multicast traffic experience high queuing latencies and significant performance penalties when running on systems designed with unicast-based NoC architectures. Various multicast mechanisms such as XY-tree multicast and path multicast have already been proposed to enhance the performance of the traditional wireline mesh NoC incorporating multicast traffic. However, even with such added features, the multihop nature of the wireline mesh NoC leads to high network latencies and thus limits the achievable system performance. In this paper, to sustain the high-bandwidth and high-throughput requirements of emerging applications, we propose the design of a wireless NoC (WiNoC) architecture incorporating necessary multicast support. By integrating congestion-aware multicast routing with network coding, the WiNoC is able to efficiently handle heavy multicast injections. For applications running with a broadcast-heavy Hammer cache coherence protocol, the proposed multicast-aware WiNoC achieves an average of 47% reduction in message latency compared with the XY-tree-based multicast-aware mesh NoC. This network level improvement translates into a 26% saving in full-system energy delay product.

Disease Diagnosis-on-a-Chip: Large Scale Networks-on-Chip based Multicore Platform for Protein Folding Analysis

Protein folding is critical for many biological processes. In this work, we propose an NoC-based multi-core platform for protein folding computation. We first identify the speedup bottleneck for applying conventional genetic algorithm on a mesh-based multi-core platform. Then, we address this computation- and communication- intensive problem while taking into account both hardware and software aspects. Specifically, we group the processing cores into islands and propose an NoC-based multicore architecture for intra- and inter-island communication. The high scalability of the proposed platform allows us to integrate from 100 to 1200 cores for the folding computation. We then propose a genetic migration algorithm to take advantage of the massive parallel platform. Our simulation results show that the proposed platform offers near-linear speedup as the number of cores increases. We also report the hardware cost in area and power based on a 100-core FPGA prototype.

Constructing compact causal mathematical models for complex dynamics

From microbial communities, human physiology to social and bio- logical/neural networks, complex interdependent systems display multi-scale spatio-temporal pa erns that are frequently classi ed as non-linear, non-Gaussian, non-ergodic, and/or fractal. Distin- guishing between the sources of nonlinearity, identifying the na- ture of fractality (space versus time) and encapsulating the non- Gaussian characteristics into dynamic causal models remains a ma- jor challenge for studying complex systems. In this paper, we pro- pose a new mathematical strategy for constructing compact yet ac- curate models of complex systems dynamics that aim to scrutinize the causal e ects and in uences by analyzing the statistics of the magnitude increments and the inter-event times of stochastic pro- cesses. We derive a framework that enables to incorporate knowl- edge about the causal dynamics of the magnitude increments and the inter-event times of stochastic processes into a multi-fractional order nonlinear partial di erential equation for the probability to nd the system in a speci c state at one time. Rather than follow- ing the current trends in nonlinear system modeling which pos- tulate speci c mathematical expressions, this mathematical frame- work enables us to connect the microscopic dependencies between the magnitude increments and the inter-event times of one stochas- tic process to other processes and justify the degree of nonlinearity. In addition, the newly presented formalism allows to investigate appropriateness of using multi-fractional order dynamical models for various complex system which was overlooked in the literature. We run extensive experiments on several sets of physiological pro- cesses and demonstrate that the derived mathematical models o er superior accuracy over state of the art techniques.

Mathematical Models and Control Algorithms for Dynamic Optimization of Multicore Platforms: A Complex Dynamics Approach

The continuous increase in integration densities contributed to a shift from Dennards scaling to a parallelization era of multi-/many-core chips. However, for multicores to rapidly percolate the application domain from consumer multimedia to high-end functionality (e.g., security, healthcare, big data), power/energy and thermal efficiency challenges must be addressed. Increased power densities can raise on-chip temperatures, which in turn decrease chip reliability and performance, and increase cooling costs. For a dependable multicore system, dynamic optimization (power / thermal management) has to rely on accurate yet low complexity workload models. Towards this end, we present a class of mathematical models that generalize prior approaches and capture their time dependence and long-range memory with minimum complexity. This modeling framework serves as the basis for defining new efficient control and prediction algorithms for hierarchical dynamic power management of future data-centers-on-a-chip.

Reliable Multi-Fractal Characterization of Weighted Complex Networks: Algorithms and Implications

Through an elegant geometrical interpretation, the multi-fractal analysis quantifies the spatial and temporal irregularities of the structural and dynamical formation of complex networks. Despite its effectiveness in unweighted networks, the multi-fractal geometry of weighted complex networks, the role of interaction intensity, the influence of the embedding metric spaces and the design of reliable estimation algorithms remain open challenges. To address these challenges, we present a set of reliable multi-fractal estimation algorithms for quantifying the structural complexity and heterogeneity of weighted complex networks. Our methodology uncovers that (i) the weights of complex networks and their underlying metric spaces play a key role in dictating the existence of multi-fractal scaling and (ii) the multi-fractal scaling can be localized in both space and scales. In addition, this multi-fractal characterization framework enables the construction of a scaling-based similarity metric and the identification of community structure of human brain connectome. The detected communities are accurately aligned with the biological brain connectivity patterns. This characterization framework has no constraint on the target network and can thus be leveraged as a basis for both structural and dynamic analysis of networks in a wide spectrum of applications.

Fundamental Challenges Toward Making the IoT a Reachable Reality: A Model-Centric Investigation

Constantly advancing integration capability is paving the way for the construction of the extremely large scale continuum of the Internet where entities or things from vastly varied domains are uniquely addressable and interacting seamlessly to form a giant networked system of systems known as the Internet-of-Things (IoT). In contrast to this visionary networked system paradigm, prior research efforts on the IoT are still very fragmented and confined to disjoint explorations of different applications, architecture, security, services, protocol, and economical domains, thus preventing design exploration and optimization from a unified and global perspective. In this context, this survey article first proposes a mathematical modeling framework that is rich in expressivity to capture IoT characteristics from a global perspective. It also sets forward a set of fundamental challenges in sensing, decentralized computation, robustness, energy efficiency, and hardware security based on the proposed modeling framework. Possible solutions are discussed to shed light on future development of the IoT system paradigm.