Improving web server performance by network aware data buffering and caching

In this paper we propose a new method of data handling for web servers. We call this method Network Aware Buffering and Caching (NABC for short). NABC facilitates reduction of data copies in web server's data sending path, by doing three things: (1) Layout the data in main memory in a way that protocol processing can be done without data copies (2) Keep a unified cache of data in kernel and ensure safe access to it by various processes and kernel and (3) Pass only the necessary meta data between processes so that bulk data handling time spent during IPC can be reduced. We realize NABC by implementing a set of system calls and an user library. The end product of the implementation is a set of APIs specifically designed for use by the web servers. We port an in house web server called SWEET, to NABC APIs and evaluate performance using a range of workloads both simulated and real. The results show a very impressive gain of 12% to 21% in throughput for static file serving and 1.6 to 4 times gain in throughput for lightweight dynamic content serving for a server using NABC APIs over the one using UNIX APIs.

Explicit Codes Uniformly Reducing Repair Bandwidth in Distributed Storage

A distributed storage setting is considered where a file of size B is to be stored across n storage nodes. A data collector should be able to reconstruct the entire data by downloading the symbols stored in any k nodes. When a node fails, it is replaced by a new node by downloading data from some of the existing nodes. The amount of download is termed as repair bandwidth. One way to implement such a system is to store one fragment of an (n, k) MDS code in each node, in which case the repair bandwidth is B. Since repair of a failed node consumes network bandwidth, codes reducing repair bandwidth are of great interest. Most of the recent work in this area focuses on reducing the repair bandwidth of a set of k nodes which store the data in uncoded form, while the reduction in the repair bandwidth of the remaining nodes is only marginal. In this paper, we present an explicit code which reduces the repair bandwidth for all the nodes to approximately B/2. To the best of our knowledge, this is the first explicit code which reduces the repair bandwidth of all the nodes for all feasible values of the system parameters.

Explicit and Optimal Codes for Distributed Storage

In the distributed storage coding problem we consider, data is stored across n nodes in a network, each capable of storing � symbols. It is required that the complete data can be reconstructed by downloading data from any k nodes. There is also the key additional requirement that a failed node be regenerated by connecting to any d nodes and downloading �symbols from each of them. Our goal is to minimize the repair bandwidth d�. In this paper we provide explicit constructions for several parameter sets of interest.

CUSUM Based Distributed Detection in WSNs

We are concerned with the situation in which a wireless sensor network is deployed in a region, for the purpose of detecting an event occurring at a random time and at a random location. The sensor nodes periodically sample their environment (e.g., for acoustic energy),process the observations (in our case, using a CUSUM-based algorithm) and send a local decision (which is binary in nature) to the fusion centre. The fusion centre collects these local decisions and uses a fusion rule to process the sensors’ local decisions and infer the state of nature, i.e., if an event has occurred or not. Our main contribution is in analyzing two local detection rules in combination with a simple fusion rule. The local detection algorithms are based on the nonparametric CUSUMprocedure from sequential statistics. We also propose two ways to operate the local detectors after an alarm. These alternatives when combined in various ways yield several approaches. Our contribution is to provide analytical techniques to calculate false alarm measures, by the use of which the local detector thresholds can be set. Simulation results are provided to evaluate the accuracy of our analysis. As an illustration we provide a design example. We also use simulations to compare the detection delays incurred in these algorithms.

Packet Reordering in Network Processors

Network processors today consist of multiple parallel processors (micro engines) with support for multiple threads to exploit packet level parallelism inherent in network workloads. With such concurrency, packet ordering at the output of the network processor cannot be guaranteed. This paper studies the effect of concurrency in network processors on packet ordering. We use a validated Petri net model of a commercial network processor, Intel IXP 2400, to determine the extent of packet reordering for IPv4 forwarding application. Our study indicates that in addition to the parallel processing in the network processor, the allocation scheme for the transmit buffer also adversely impacts packet ordering. In particular, our results reveal that these packet reordering results in a packet retransmission rate of up to 61%. We explore different transmit buffer allocation schemes namely, contiguous, strided, local, and global which reduces the packet retransmission to 24%. We propose an alternative scheme, packet sort, which guarantees complete packet ordering while achieving a throughput of 2.5 Gbps. Further, packet sort outperforms the in-built packet ordering schemes in the IXP processor by up to 35%.

Co-ordinate interleaved distributed space-time coding for two-antenna relay networks

Distributed space time coding for wireless relay networks where the source, the destination and the relays have multiple antennas have been studied by Jing and Hassibi. In this set up, the transmit and the receive signals at different antennas of the same relay are processed and designed independently, even though the antennas are colocated. In this paper, a wireless relay network with single antenna at the source and the destination and two antennas at each of the R relays is considered. In the first phase of the two-phase transmission model, a T -length complex vector is transmitted from the source to all the relays. At each relay, the inphase and quadrature component vectors of the received complex vectors at the two antennas are interleaved before processing them. After processing, in the second phase, a T x 2R matrix codeword is transmitted to the destination. The collection of all such codewords is called Co-ordinate interleaved distributed space-time code (CIDSTC). Compared to the scheme proposed by Jing-Hassibi, for T ges AR, it is shown that while both the schemes give the same asymptotic diversity gain, the CIDSTC scheme gives additional asymptotic coding gain as well and that too at the cost of negligible increase in the processing complexity at the relays.

A Distributed Approach to Interference Cancellation

In this paper, we propose and analyze a novel idea of performing interference cancellation (IC) in a distributed/cooperative manner, with a motivation to provide multiuser detection (MUD) benefit to nodes that have only a single user detection capability. In the proposed distributed interference cancellation (DIC) scheme, during phase-1 of transmission, an MUD capable cooperating relay node estimates all the sender nodes' bits through multistage interference cancellation. These estimated bits are then sent by the relay node on orthogonal tones in phase-2 of transmission. The destination nodes receive these bit estimates and use them for interference estimation/cancellation, thus achieving IC benefit in a distributed manner. For this DIC scheme, we analytically derive an exact expression for the bit error rate (BER) in a basic five-node network (two source-destination node pairs and a cooperating relay node) on AWGN channels. Analytical BER results are shown to match with simulation results. For more general system scenarios, including more than two source-destination pairs and fading channels without and with space-time coding, we present simulation results to establish the potential for improved performance in the proposed distributed approach to interference cancellation. We also present a linear version of the proposed DIC.

On four-group ML decodable distributed space time codes for cooperative communication

A construction of a new family of distributed space time codes (DSTCs) having full diversity and low Maximum Likelihood (ML) decoding complexity is provided for the two phase based cooperative diversity protocols of Jing-Hassibi and the recently proposed Generalized Non-orthogonal Amplify and Forward (GNAF) protocol of Rajan et al. The salient feature of the proposed DSTCs is that they satisfy the extra constraints imposed by the protocols and are also four-group ML decodable which leads to significant reduction in ML decoding complexity compared to all existing DSTC constructions. Moreover these codes have uniform distribution of power among the relays as well as in time. Also, simulations results indicate that these codes perform better in comparison with the only known DSTC with the same rate and decoding complexity, namely the Coordinate Interleaved Orthogonal Design (CIOD). Furthermore, they perform very close to DSTCs from field extensions which have same rate but higher decoding complexity.

Training-Symbol Embedded, High-Rate, Single-Symbol ML-Decodable, Distributed STBCs for Relay Networks

Distributed space-time block codes (DSTBCs) from complex orthogonal designs (CODs) (both square and nonsquare), coordinate interleaved orthogonal designs (CIODs), and Clifford unitary weight designs (CUWDs) are known to lose their single-symbol ML decodable (SSD) property when used in two-hop wireless relay networks using amplify and forward protocol. For such networks, in this paper, three new classes of high rate, training-symbol embedded (TSE) SSD DSTBCs are constructed: TSE-CODs, TSE-CIODs, and TSE-CUWDs. The proposed codes include the training symbols inside the structure of the code which is shown to be the key point to obtain the SSD property along with the channel estimation capability. TSE-CODs are shown to offer full-diversity for arbitrary complex constellations and the constellations for which TSE-CIODs and TSE-CUWDs offer full-diversity are characterized. It is shown that DSTBCs from nonsquare TSE-CODs provide better rates (in symbols per channel use) when compared to the known SSD DSTBCs for relay networks. Important from the practical point of view, the proposed DSTBCs do not contain any zeros in their codewords and as a result, antennas of the relay nodes do not undergo a sequence of switch on/off transitions within every codeword, and, thus, avoid the antenna switching problem.

Optimal Exact-Regenerating Codes for Distributed Storage at the MSR and MBR Points via a Product-Matrix Construction

Regenerating codes are a class of distributed storage codes that allow for efficient repair of failed nodes, as compared to traditional erasure codes. An [n, k, d] regenerating code permits the data to be recovered by connecting to any k of the n nodes in the network, while requiring that a failed node be repaired by connecting to any d nodes. The amount of data downloaded for repair is typically much smaller than the size of the source data. Previous constructions of exact-regenerating codes have been confined to the case n = d + 1. In this paper, we present optimal, explicit constructions of (a) Minimum Bandwidth Regenerating (MBR) codes for all values of [n, k, d] and (b) Minimum Storage Regenerating (MSR) codes for all [n, k, d >= 2k - 2], using a new product-matrix framework. The product-matrix framework is also shown to significantly simplify system operation. To the best of our knowledge, these are the first constructions of exact-regenerating codes that allow the number n of nodes in the network, to be chosen independent of the other parameters. The paper also contains a simpler description, in the product-matrix framework, of a previously constructed MSR code with [n = d + 1, k, d >= 2k - 1].

An Empirical Analysis of Heterogeneity in IEEE 802.11 MAC Protocol Implementations and its Implications

Wireless LAN (WLAN) market consists of IEEE 802.11 MAC standard conformant devices (e.g., access points (APs), client adapters) from multiple vendors. Certain third party certifications such as those specified by the Wi-Fi alliance have been widely used by vendors to ensure basic conformance to the 802.11 standard, thus leading to the expectation that the available devices exhibit identical MAC level behavior. In this paper, however, we present what we believe to be the first ever set of experimental results that highlight the fact that WLAN devices from different vendors in the market can have heterogeneous MAC level behavior. Specifically, we demonstrate with examples and data that in certain cases, devices may not be conformant with the 802.11 standard while in other cases, they may differ in significant details that are not a part of mandatory specifications of the standard. We argue that heterogeneous MAC implementations can adversely impact WLAN operations leading to unfair bandwidth allocation, potential break-down of related MAC functionality and difficulties in provisioning the capacity of a WLAN. However, on the positive side, MAC level heterogeneity can be useful in applications such as vendor/model level device fingerprinting.

Exploiting Programmable Network Interfaces for Parallel Query Execution in Workstation Clusters

Workstation clusters equipped with high performance interconnect having programmable network processors facilitate interesting opportunities to enhance the performance of parallel application run on them. In this paper, we propose schemes where certain application level processing in parallel database query execution is performed on the network processor. We evaluate the performance of TPC-H queries executing on a high end cluster where all tuple processing is done on the host processor, using a timed Petri net model, and find that tuple processing costs on the host processor dominate the execution time. These results are validated using a small cluster. We therefore propose 4 schemes where certain tuple processing activity is offloaded to the network processor. The first 2 schemes offload the tuple splitting activity - computation to identify the node on which to process the tuples, resulting in an execution time speedup of 1.09 relative to the base scheme, but with I/O bus becoming the bottleneck resource. In the 3rd scheme in addition to offloading tuple processing activity, the disk and network interface are combined to avoid the I/O bus bottleneck, which results in speedups up to 1.16, but with high host processor utilization. Our 4th scheme where the network processor also performs apart of join operation along with the host processor, gives a speedup of 1.47 along with balanced system resource utilizations. Further we observe that the proposed schemes perform equally well even in a scaled architecture i.e., when the number of processors is increased from 2 to 64

An adaptive MAC protocol in MANETs: a democratic approach

Mobile ad-hoc networks (MANETs) have recently drawn significant research attention since they offer unique benefits and versatility with respect to bandwidth spatial reuse, intrinsic fault tolerance, and low-cost rapid deployment. This paper addresses the issue of delay sensitive realtime data transport in these type of networks. An effective QoS mechanism is thereby required for the speedy transport of the realtime data. QoS issue in MANET is an open-end problem. Various QoS measures are incorporated in the upperlayers of the network, but a few techniques addresses QoS techniques in the MAC layer. There are quite a few QoS techniques in the MAC layer for the infrastructure based wireless network. The goal and the challenge is to achieve a QoS delivery and a priority access to the real time traffic in adhoc wireless environment, while maintaining democracy in the resource allocation. We propose a MAC layer protocol called "FCP based FAMA protocol", which allocates the channel resources to the needy in a more democratic way, by examining the requirements, malicious behavior and genuineness of the request. We have simulated both the FAMA as well as FCP based FAMA and tested in various MANET conditions. Simulated results have clearly shown a performance improvement in the channel utilization and a decrease in the delay parameters in the later case. Our new protocol outperforms the other QoS aware MAC layer protocols.

Distributed load adaptive scheduling for throughput and delay guarantees in fading wireless LANs

We consider a problem of providing mean delay and average throughput guarantees in random access fading wireless channels using CSMA/CA algorithm. This problem becomes much more challenging when the scheduling is distributed as is the case in a typical local area wireless network. We model the CSMA network using a novel queueing network based approach. The optimal throughput per device and throughput optimal policy in an M device network is obtained. We provide a simple contention control algorithm that adapts the attempt probability based on the network load and obtain bounds for the packet transmission delay. The information we make use of is the number of devices in the network and the queue length (delayed) at each device. The proposed algorithms stay within the requirements of the IEEE 802.11 standard.

Fracture analysis of concrete using singular fractal functions with lattice beam network and confirmation with acoustic emission study

In the present study singular fractal functions (SFF) were used to generate stress-strain plots for quasibrittle material like concrete and cement mortar and subsequently stress-strain plot of cement mortar obtained using SFF was used for modeling fracture process in concrete. The fracture surface of concrete is rough and irregular. The fracture surface of concrete is affected by the concrete's microstructure that is influenced by water cement ratio, grade of cement and type of aggregate 11-41. Also the macrostructural properties such as the size and shape of the specimen, the initial notch length and the rate of loading contribute to the shape of the fracture surface of concrete. It is known that concrete is a heterogeneous and quasi-brittle material containing micro-defects and its mechanical properties strongly relate to the presence of micro-pores and micro-cracks in concrete 11-41. The damage in concrete is believed to be mainly due to initiation and development of micro-defects with irregularity and fractal characteristics. However, repeated observations at various magnifications also reveal a variety of additional structures that fall between the `micro' and the `macro' and have not yet been described satisfactorily in a systematic manner [1-11,15-17]. The concept of singular fractal functions by Mosolov was used to generate stress-strain plot of cement concrete, cement mortar and subsequently the stress-strain plot of cement mortar was used in two-dimensional lattice model [28]. A two-dimensional lattice model was used to study concrete fracture by considering softening of matrix (cement mortar). The results obtained from simulations with lattice model show softening behavior of concrete and fairly agrees with the experimental results. The number of fractured elements are compared with the acoustic emission (AE) hits. The trend in the cumulative fractured beam elements in the lattice fracture simulation reasonably reflected the trend in the recorded AE measurements. In other words, the pattern in which AE hits were distributed around the notch has the same trend as that of the fractured elements around the notch which is in support of lattice model. (C) 2011 Elsevier Ltd. All rights reserved.