INFOCOM 2015 Papers

This paper studies the task of setting up ad hoc wireless networks. In such networks, it is often the case that nodes become active at different times, without coordination or knowledge about network topology. We consider the following tasks: wake-up, clock synchronization, leader election, and multimessage broadcast. We show how to achieve these goals in scalable O(D polylog(n)) time. As a tool we define and give a solution to a quasi-backbone problem, which aims to set up transmission probabilities at nodes in a way that they can be efficiently used to solve other tasks. Our results are obtained by minimalistic algorithms, which do not require power control or carrier sensing capabilities, use very small energy, local computation and memory. Moreover, unlike many previous work, they remain scalable even if the network is not highly connected.

In RFID systems, the grouping problem is to efficiently group all tags according to a given partition such that tags in the same group will have the same group ID. Unlike previous research on the unicast transmission from a reader to a tag, grouping provides a fundamental mechanism for efficient multicast transmissions and aggregate queries in large RFID-enabled applications. A message can be transmitted to a group of m tags simultaneously in multicast, which improves the efficiency by m times when comparing with unicast. We study fast grouping protocols in large RFID systems. To the best of our knowledge, it is the first attempt to tackle this practically important yet uninvestigated problem. We start with a straightforward solution called the Enhanced Polling Grouping (EPG) protocol. We then propose a time-efficient FIltering Grouping (FIG) protocol that uses Bloom filters to remove the costly ID transmissions. We point out the limitation of the Bloom-filter based solution due to its intrinsic false positive problem, which leads to our final ConCurrent Grouping (CCG) protocol. With a drastically different design, CCG is able to outperform FIG by exploiting collisions to inform multiple tags of their group ID simultaneously and by removing any wasteful slots in its frame-based execution. Simulation results demonstrate that our best protocol CCG can reduce the execution time by a factor of 11 when comparing with a baseline polling protocol.

This paper considers the end-to-end scheduling for all-optical data center networks with zero in-network buffer and non-negligible reconfiguration delay. It is known that in the regime where the scheduling reconfiguration delay is non-negligible, the rate of schedule reconfiguration should be limited in such a way as to minimize the impact of reduced duty-cycles and to ensure bounded delay. However, when the scheduling rate is restricted, the existing literature also tends to restrict the rate of monitoring and decision processes. We first present a framework for scheduling with reconfiguration delay that decouples the rate of scheduling from the rate of monitoring. Under this framework, we then present two scheduling algorithms for switches with reconfiguration delay, both based on the well-known MaxWeight scheduling policy. The first one is the Periodic MaxWeight (PMW), which is simpler in computation, but requires prior knowledge of traffic load. The other is the Adaptive MaxWeight (AMW), which, in contrast, requires no prior knowledge. We show the stability condition for both algorithms and evaluate their delay performance through simulations.

Energy efficiency of server farms is an important design consideration of data centers. One effective approach is to optimize energy consumption by controlling carried load on the networked servers. In this paper, we propose a robust heuristic policy for job assignment in a server farm, aiming to improve the energy efficiency by maximizing the ratio of the long-run average throughput to the expected energy consumption. Our model of the server farm considers parallel processor-sharing queues with finite buffer sizes, heterogeneous server speeds, and an arbitrary energy consumption function. We devise the new energy-efficient (EE) policy in a way that the state distribution of the system depends on the service requirement distribution only through the mean. We show that the state-of-the-art slowest server first (SSF) policy can be obtained as a special case of EE and both policies have the same computational complexity. We provide a rigorous analysis of EE and derive conditions under which EE is guaranteed to outperform SSF in terms of energy efficiency. Extensive numerical results are presented and demonstrate that, in comparison with SSF, EE yields a consistently better system throughput and yet improves the energy efficiency by up to 70%.

We consider the task of assigning time slots on a user-dependent and time-varying wireless channel. This scheduling problem occurs in cellular networks due to the presence of channel fading and user mobility. We introduce a simple notion of global fairness, where each of n users is guaranteed a 1/(n + ε) fraction of its total possible throughput, for some approximation parameter ε ≥ 0, and study its limitations from theoretical and experimental perspectives. We formally prove that a slight modification of the standard proportional fair algorithm satisfies the global fairness constraint. To the best of our knowledge, this is the first formal analysis providing global fairness property to the channel in any execution and any channel conditions. As confirmed by our simulations, our global fairness constraint is in fact satisfied by a wide class of algorithms. Our framework allows optimization of an arbitrary metric subject to the global fairness constraint. In particular, we have analyzed a variant of the provably fair algorithm that optimizes the total throughput. It turned out that the channel utilization of this algorithm is significantly better than that of the classical Proportional Fair algorithm.

Data harvesting using mobile data ferries has recently emerged as a promising alternative to the traditional multi-hop transmission paradigm. The use of data ferries can significantly reduce energy consumption at sensor nodes and increase network lifetime. However, it usually incurs longer data delivery latency as the data ferry needs to travel through the network to collect data, during which some delay-sensitive data may become obsolete. Therefore, optimizing the trajectory of the data ferry with data delivery latency bound is important for this approach to be effective in practice. To address this problem, we formally define the time-constrained data harvesting problem, which seeks an optimal data harvesting path in a network to collect as much data as possible within a time duration. We first characterise the performance bound given by the optimal data harvesting algorithm and show that the optimal algorithm significantly outperforms the random algorithm, especially when network scales. Motivated by the theoretical analysis and proving the NP-completeness of the time-constrained data harvesting problem, we then devise polynomial-time approximation schemes (PTAS) and mathematically prove the output being a constant-factor approximation of the optimal solution.

The benefits of network coding on multicast in traditional multi-hop wireless networks have already been demonstrated in previous works. However, most existing approaches cannot be directly applied to multi-hop cognitive radio networks (CRNs), given the unpredictable primary user occupancy on licensed channels. Specifically, due to the unpredictable occupancy, the channel's bandwidth is uncertain and thus the capacity of the link using this channel is also uncertain, which may result in severe throughput loss. In this paper, we study the problem of network coding-based multicast in multi-hop CRNs considering the uncertain spectrum availability. To capture the uncertainty of spectrum availability, we first formulate our problem as a chance-constrained program. Given the computationally intractability of the above program, we transform the original problem into a tractable convex optimization problem, through appropriate Bernstein approximation together with relaxation on link scheduling. We further leverage Lagrangian relaxation-based optimization techniques to propose an efficient distributed algorithm for the original problem. Extensive simulation results show that, the proposed algorithm achieves higher multicast rates, compared to a state-of-the-art non-network coding algorithm in multi-hop CRNs, and a conservative robust algorithm that treats the link capacity as a constant value in the optimization.

Solving large-scale linear systems of equations (LSEs) is one of the most common and fundamental problems in big data. But such problems are often too expensive to solve for resource-limited users. Cloud computing has been proposed as a timely, efficient, and cost-effective way of solving such computing tasks. Nevertheless, one critical concern in cloud computing is data privacy. To be more prominent, in many cases, clients's LSEs contain private data that should remain hidden from the cloud for ethical, legal, or security reasons. Many previous works on secure outsourcing of LSEs have high computational complexity. More importantly, they share a common serious problem, i.e., a huge number of external memory I/O operations. This problem has been largely neglected in the past, but in fact is of particular importance and may eventually render those outsourcing schemes impractical. In this paper, we develop an efficient and practical secure outsourcing algorithm for solving large-scale LSEs, which has both low computational complexity and low memory I/O complexity and can protect clients' privacy well. We implement our algorithm on a real-world cloud server and a laptop. We find that the proposed algorithm offers significant time savings for the client (up to 65%) compared to previous algorithms.

In IaaS cloud environments, peak memory demand caused by hotspot applications in Virtual Machine (VM) often results in performance degradation within and outside of this VM. Some solutions such as host swapping and ballooning for memory consolidation and overcommitment have been proposed. These solutions, however, have no help for addressing guest swapping issues inside VM. Even though host holds sufficient memory pages, guest OS is unable to utilize free pages in host directly due to the semantic gap between VMM and it. Our goal is to alleviate the performance degradation by decreasing disk I/O operations generated by guest swapping. Based on the insight analysis of behavioral features of guest swapping, we design HybridSwap, a distributed scalable framework which organize surplus memory in all hosts within data center into virtual pools for swapping. This framework builds up a synthetic swapping mechanism in a peer-to-peer way, which VM can adaptively choose suitable pools for swapping. We implement the prototype of HybridSwap and evaluate it with different benchmarks. The results demonstrate that our solution has the ability to promote the guest swapping efficiency indeed. Even in some cases, it shows 2-5 times of performance promotion compared with the baseline setup.

Current traffic engineering in SDN mostly focuses on unicast. By contrast, compared with individual unicast, multicast can effectively reduce network resources consumption to serve multiple clients jointly. Since many important applications require reliable transmissions, it is envisaged that reliable multicast plays a crucial role when an SDN operator plans to provide multicast services. However, the shortest-path tree (SPT) adopted in current Internet is not bandwidth-efficient, while the Steiner tree (ST) in Graph Theory is not designed to support reliable transmissions since the selection of recovery nodes is not examined. In this paper, therefore, we propose a new reliable multicast tree for SDN, named Recover-aware Steiner Tree (RST). The goal of RST is to minimize both tree and recovery costs, while finding an RST is very challenging. We prove that the RST problem is NP-Hard and inapproximable within k, which is the number of destination nodes. Thus, we design an approximate algorithm, called Recover Aware Edge Reduction Algorithm (RAERA), to solve the problem. The simulation results on real networks and large synthetic networks, together with the experiment on our SDN testbed with real YouTube traffic, all manifest that RST outperforms both SPT and ST. Also, the implementation of RAERA in SDN controllers shows that an RST can be returned within a few seconds and thereby is practical for SDN networks.

We develop a unifying framework to obtain efficient index policies for restless multi-armed bandit problems with birth-and-death state evolution. This is a broad class of stochastic resource allocation problems whose objective is to determine efficient policies to share resources among competing projects. In a seminal work, Whittle developed a methodology to derive well-performing (Whittle's) index policies that are obtained by solving a relaxed version of the original problem. Our first main contribution is the derivation of a closed-form expression for Whittle's index as a function of the steady-state probabilities. It can be efficiently calculated, however, it requires several technical conditions to be verified, and in addition, it does not provide qualitative insights into Whittle's index. We therefore formulate a fluid version of the relaxed optimization problem and in our second main contribution we develop a fluid index policy. The latter does provide qualitative insights and is close to Whittle's index. The applicability of our approach is illustrated by two important problems: optimal class selection and optimal load balancing. Allowing state-dependent capacities we can model important phenomena: e.g. power-aware server-farms and opportunistic scheduling in wireless systems. Numerical simulations show that Whittle's index and our fluid index policy are both nearly optimal.

Modern network elements are increasingly required to deal with heterogeneous traffic. Recent works consider processing policies for buffers that hold packets with different processing requirement (number of processing cycles needed before a packet can be transmitted out) but uniform value, aiming to maximize the throughput, i.e., the number of transmitted packets. Other developments deal with packets of varying value but uniform processing requirement (each packet requires one processing cycle); the objective here is to maximize the total transmitted value. In this work, we consider a more general problem, combining packets with both nonuniform processing and nonuniform values in the same queue. We study the properties of various processing orders in this setting. We show that in the general case natural processing policies have poor performance guarantees, with linear lower bounds on their competitive ratio. Moreover, we show an adversarial lower bound that holds for every online policy. On the positive side, in the special case when only two different values are allowed, 1 and V, we present a policy that achieves competitive ratio (1 + W+2/V), where W is the maximal number of required processing cycles. We also consider copying costs during admission.

DDoS attacks have been a persistent threat to network availability for many years. Most of the existing mitigation techniques attempt to protect against DDoS by filtering out attack traffic. However, as critical network resources are usually static, adversaries are able to bypass filtering by sending stealthy low traffic from large number of bots that mimic benign traffic behavior. Sophisticated stealthy attacks on critical links can cause a devastating effect such as partitioning domains and networks. In this paper, we propose to defend against DDoS attacks by proactively changing the footprint of critical resources in an unpredictable fashion to invalidate an adversary's knowledge and plan of attack against critical network resources. Our present approach employs virtual networks (VNs) to dynamically reallocate network resources using VN placement and offers constant VN migration to new resources. Our approach has two components: (1) a correct-by-construction VN migration planning that significantly increases the uncertainty about critical links of multiple VNs while preserving the VN placement properties, and (2) an efficient VN migration mechanism that identifies the appropriate configuration sequence to enable node migration while maintaining the network integrity (e.g., avoiding session disconnection). We formulate and implement this framework using SMT logic. We also demonstrate the effectiveness of our implemented framework on both PlanetLab and Mininet-based experimentations.

It is well-known that maximum weight scheduling, with link weights which are either functions of queue lengths or the ages of the Head-of-Line (HoL) packets in each queue, maximizes the throughput region of wireless networks with persistent flows. In particular, with only persistent flows, it does not matter for throughput optimality whether one uses queue lengths or HoL ages as weights. In this paper, we show the following interesting result: when some flows in the network are dynamic (i.e., they arrive and depart from the network and are not persistent), then HoL-age-based scheduling algorithms are throughput-optimal while it has previously been shown that queue-length-based algorithms are not. This reveals that, age-based algorithms are universal in the sense that their throughput optimality does not depend on whether the arriving traffic is persistent or not. We also present a distributed implementation of the proposed age-based algorithm using CSMA techniques, where each flow only knows its own age and carrier sensing information. Finally, we support our analytical results through simulations. The proof of throughput optimality may be interesting in its own right: it uses a novel Lyapunov function which is the sum of the ages of all the packets in the network.

To improve spectrum utilization, cognitive radio (CR) is introduced to detect and exploit available spectrum resources autonomously. The flexible spectrum use imposes special challenges on broadcast because different CR devices may detect different available spectrum fragments at different locations. The sender and the receivers have to agree on spectrum fragments that will be used for broadcast. There may not exist a common spectrum fragment that is available to all receivers. Most existing work assumes that a device works only in a single channel and thus the sender has to broadcast multiple times in different channels to reach all receivers. The broadcast problem is studied as a channel rendezvous and minimum latency scheduling problem. Recent spectrum-agile designs have enabled a device to utilize partially occupied spectrum. We thus view a wideband channel as an aggregation of multiple narrow channels that can be evaluated independently. A Spectrum Fragment Agile Broadcast (SFAB) scheme is introduced in this paper to support efficient broadcast on fragmented spectrum. It aims at achieving spectrum agreement between the transmitter and the receivers efficiently and maximizing the channel width used for broadcast regardless of the spectrum availability differences at receivers. We validate the effectiveness of SFAB through implementation on the GNU Radio / USRP platform and use ns-2 simulations to evaluate the performance in large deployments.

Live migration is a key technique for virtual machine (VM) management in data center networks, which enables flexibility in resource optimization, fault tolerance, and load balancing. Despite its usefulness, the live migration still introduces performance degradations during the migration process. Thus, there has been continuous efforts in reducing the migration time in order to minimize the impact. From the network's perspective, the migration time is determined by the amount of data to be migrated and the available bandwidth used for such transfer. In this paper, we examine the problem of how to schedule the migrations and how to allocate network resources for migration when multiple VMs need to be migrated at the same time. We consider the problem in the Software-defined Network (SDN) context since it provides flexible control on routing. More specifically, we propose a method that computes the optimal migration sequence and network bandwidth used for each migration. We formulate this problem as a mixed integer programming, which is NP-hard. To make it computationally feasible for large scale data centers, we propose an approximation scheme via linear approximation plus fully polynomial time approximation, and obtain its theoretical performance bound. Through extensive simulations, we demonstrate that our fully polynomial time approximation (FPTA) algorithm has a good performance compared with the optimal solution and two state-of-the-art algorithms. That is, our proposed FPTA algorithm approaches to the optimal solution with less than 10% variation and much less computation time. Meanwhile, it reduces the total migration time and the service downtime by up to 40% and 20% compared with the state-of-the-art algorithms, respectively.

In many networks the operator is faced with nodes that report a potentially important phenomenon such as failures, illnesses, and viruses. The operator is faced with the question: Is it spreading over the network, or simply occurring at random? We seek to answer this question from highly noisy and incomplete data, where at a single point in time we are given a possibly very noisy subset of the infected population (including false positives and negatives). While previous work has focused on uniform spreading rates for the infection, heterogeneous graphs with unequal edge weights are more faithful models of reality. Critically, the network structure may not be fully known and modeling epidemic spread on unknown graphs relies on non-homogeneous edge (spreading) weights. Such heterogeneous graphs pose considerable challenges, requiring both algorithmic and analytical development. We develop an algorithm that can distinguish between a spreading phenomenon and a randomly occurring phenomenon while using only local information and not knowing the complete network topology and the weights. Further, we show that this algorithm can succeed even in the presence of noise, false positives and unknown graph edges.

To meet the demand of more intelligent automation services on smartphone, more and more applications are developed based on users' emotion and personality. It has been a consensus that a relationship exists between personal emotions and usage pattern of smartphone. Most of existing work studies this relationship by learning manually labeled samples collected from smartphone users. The manual labeling process, however, is time-consuming, labor-intensive and money-consuming. To address this issue, we propose iSelf, a system which provides a general service of automatic detection for user's emotions in cold-start conditions with smartphone. Using transfer learning technology, iSelf achieves high accuracy given only a few labeled samples. We also develop a hybrid public/personal inference engine and validation system, so as to make iSelf maintain continuous update. Through extensive experiments, the inferring accuracy is tested about 75% and can be improved increasingly through validation and update.

On-demand resource provisioning in cloud computing provides tailor-made resource packages (typically in the form of VMs) to meet users' demands. Public clouds nowadays provide more and more elaborated types of VMs, but have yet to offer the most flexible dynamic VM assembly, which is partly due to the lack of a mature mechanism for pricing tailor-made VMs on the spot. This work proposes an efficient randomized auction mechanism based on a novel application of smoothed analysis and randomized reduction, for dynamic VM provisioning and pricing in geo-distributed cloud data centers. This auction, to the best of our knowledge, is the first one in literature that achieves (i) truthfulness in expectation, (ii) polynomial running time in expectation, and (iii) (1 − ε)-optimal social welfare in expectation for resource allocation, where e can be arbitrarily close to 0. Our mechanism consists of three modules: (1) an exact algorithm to solve the NP-hard social welfare maximization problem, which runs in polynomial time in expectation, (2) a perturbation-based randomized resource allocation scheme which produces a VM provisioning solution that is (1 − ε)-optimal and (3) an auction mechanism that applies the perturbation-based scheme for dynamic VM provisioning and prices the customized VMs using a randomized VCG payment, with a guarantee in truthfulness in expectation. We validate the efficacy of the mechanism through careful theoretical analysis and trace-driven simulations.1