Universities worldwide are experiencing a surge in enrolments, therefore campus estate managers are seeking continuous data on attendance patterns so as to optimize the usage of classroom space. While prior works have measured room occupancy via hardware sensor instrumentation, in this paper we explore the use of pervasive WiFi infrastructure for estimating attendance. In a dense campus environment, WiFi connectivity counts are poor estimators of room occupancy since they are polluted by adjoining rooms, outdoor walkways, and network load balancing. The main contribution of this work is to develop new ways to distinguish and filter out WiFi-connected users outside of the lecture room of interest, and feed such data to a regression analyser to estimate room occupancy. We evaluate our technique across lecture theatres of varying size in our campus, and show that their accuracy approaches that of hardware sensors without incurring cost and effort of installing and maintaining them.

Residential broadband consumption is growing rapidly, increasing the gap between Internet service provider (ISP) costs and revenues. Meanwhile, proliferation of Internet-enabled devices is congesting access networks, degrading end-user experience, and affecting content provider monetization. In this paper, we propose a new model whereby the content provider explicitly signals fast- and slow-lane requirements to the ISP on a per-flow basis, using open APIs supported through software defined networking (SDN). Our first contribution is to develop an architecture that supports this model, presenting arguments on why this benefits consumers (better user experience), ISPs (two-sided revenue), and content providers (fine-grained control over peering arrangement). Our second contribution is to evaluate our proposal using a real trace of over 10 million flows to show that video flow quality degradation can be nearly eliminated by the use of dynamic fast-lanes, and web-page load times can be hugely improved by the use of slow-lanes for bulk transfers. Our third contribution is to develop a fully functional prototype of our system using open-source SDN components (Openflow switches and POX controller modules) and instrumented video/file-transfer servers to demonstrate the feasibility and performance benefits of our approach. Our proposal is a first step towards the long-term goal of realizing open and agile access network service quality management that is acceptable to users, ISPs, and content providers alike.

Current Content Delivery Networks (CDNs) primarily utilize the Domain Name System (DNS) to direct users towards optimal content replica servers. Typically, these CDNs will leverage specialized DNS servers to perform load balancing, by providing different users with different IP endpoints for a particular hostname. However, there typically exists some form of DNS cache between the user and the authoritative DNS for a particular web service (Examples include web browsers, operating systems, and home routers). While these DNS caches typically improve user experience and network performance by reducing the number of redundant DNS requests, there exist scenarios where the presence of a DNS cache can actually harm network performance and user experience. For example: if a server or datacentre were to suddenly become unavailable, end users may become `stuck' trying to access the old IP address contained in their DNS caches, even though the authoritative DNS server may have updated it's records accordingly. In order to address this issue, we propose a system that enables network operators to transparently redirect users towards optimal servers during times of network congestion or high server load. By leveraging SDN and NFV, we are able to implement this system over a variety of network infrastructures, and to various levels of scale. Our initial tests indicate that our methodology has the capability to vastly improve user experience by several orders of magnitude during times of high network load, over a variety of network conditions. Our system is application-layer agnostic, and is compatible with various networking protocols.

Residential broadband consumption is growing rapidly, increasing the gap between ISP costs and revenues. Meanwhile, proliferation of Internet-enabled devices is congesting access networks, frustrating end-users and content providers. We propose that ISPs virtualize access infrastructure, using open APIs supported through SDN, to enable dynamic and controlled sharing amongst user streams. Content providers can programmatically provision capacity to user devices to ensure quality of experience, users can match the degree of virtualization to their usage pattern, and ISPs can realize per-stream revenues by slicing their network resources. Using video streaming and bulk transfers as examples, we develop an architecture that specifies the interfaces between the ISP, content provider, and user. We propose an algorithm for optimally allocating network resources, leveraging bulk transfer time elasticity and access path space diversity. Simulations using real traces show that virtualization can reduce video degradation by over 50%, for little extra bulk transfer delay. Lastly, we prototype our system and validate it in a test-bed with real video streaming and file transfers. Our proposal is a first step towards the long-term goal of realizing open and agile access network service quality management that is acceptable to users, ISPs and content providers alike.

Video already constitutes more than half of Internet traffic, and is expected to expand to over 80% of Internet traffic by 2016. The dominance of video traffic justifies router features that optimise performance when video loads are high, in particular features that can deal with the congestion that video can create. In this paper, we exploit a characteristic of video, namely that video streams consist of recurring patterns of packets of varying importance, to create a new Deferred Discard queue management scheme. We simulated different queue management schemes with four test videos and measured their performance in terms of the average picture signal to noise ratio (PSNR). In the range of PSNRs from 18.5 for best-effort service to 33.1 for ideal service, Differentiated Services (Diffserv) provided a PSNR of 22.1, while Deferred Discard improved the PSNR to 26.8 when used alone, and to 27.9 when combined with Diffserv.

This paper examines the spatial correlation of packet loss events in IEEE 802.11 wireless networks for broadcast communications. We discuss limitations of previously used metrics to measure spatial loss correlation and show that the entropy correlation coefficient, which is based on the mutual information concept of Information Theory, can overcome these limitations. In our experiments, we find that the packet losses among closely located receivers are highly correlated when signal strength is high, but the correlation decreases with decreasing signal strength. The losses become totally independent once the signal strength falls below a threshold. As a first step to model the relationship between spatial loss correlation and signal strength, we find that the relationship can be approximated as a Gaussian cumulative distribution function.

Before vehicle-to-vehicle-communication-based road safety becomes a reality, one of the challenges that needs to be resolved is efficient and timely recovery of safety packets that may be lost due to reception failure. Various types of retransmission-based recovery protocols have been proposed in the literature to address this problem. Although, retransmission can help recover some of the lost packets, it consumes significant bandwidth. In this paper, we apply the concept of network coding to achieve efficient loss recovery by combining multiple packets that were transmitted by different vehicles into a single retransmission. Through an analytical study of coding gain, we show how to optimize the performance of our proposed network coding based recovery scheme. Using simulation, we demonstrate that the use of network coding improves the bandwidth efficiency of the recovery protocols, which in turn improves the performance of vehicular safety communication.

Many distributed hash tables (DHTs) resolve lookups in O(\log n) hops, where n is the number of nodes. One-hop DHTs give lower lookup latencies and lower lookup failure rates. However, it is hard to maintain large, wide-area one-hop topologies. We contribute aecast, a new topology dissemination algorithm for one-hop DHTs. It avoids expensive repair mechanisms and critical points of failure in existing one-hop DHTs. When a node discovers by anti-entropy that it has missed a topology update, it initiates "controlled flooding,rdquo sending the update to nodes in the multicast tree that also missed the update. We compare aecast with a widely cited epidemic multicasting algorithm, pbcast, by analysis and simulation. Aecast gives at least fivefold fewer out-of-date nodes on average within one round of a topology update. We support it with a fault-tolerant topology agreement protocol, so that only legitimate topology changes propagate throughout the overlay. Consequently, we argue that one-hop DHTs deserve greater attention for Internet applications in which reasonably reliable nodes carry high lookup loads.

Overlay networks can be used to find working paths when direct underlay paths are anomalously slow, e.g. because of a network fault. Overlay paths should not use links that are involved in a fault, so choosing which overlay path to use often requires path monitoring, which introduces an overhead. By using a routing matrix ‘M’ to define which links are used in each path, and sorting the matrix according to the degree of independence of paths, we can choose a subset of paths to monitor, and so reduce overheads. The performance metrics of the unmonitored paths are then predicted based on information inferred from the monitored. Previous work has shown how statistical prediction errors can occur, even in small networks (11 nodes, 110 paths), when monitoring fewer paths than the rank of the matrix. This paper extends previous work by showing that collinear relationships between variables in paths of larger routing matrices in networks with tens of nodes and 1–2 000 paths are a large source of errors in larger networks. We show that mitigation of such errors leads to improved path metric prediction and anomaly detection.

While email is fairly reliable, messages do get lost, e.g. due to the store-and-forward nature of e-mail, which passes responsibility for delivering email from one mail server to another. Some end users are more sensitive to email unreliability than others, and seek tools with which they can detect and recover from email failures. This paper describes four such tools: An SMTP traceroute tool for determining which servers an email passes through in order to focus troubleshooting, authentication of delivery notifications so that users can focus on genuine notifications rather than those resulting from spoofing, tracking which messages have received replies (others warrant closer scrutiny for possible loss) and linking of messages sent to mailing lists to enable receivers to detect lost intermediate messages. While our implementation is based in an end-userpsilas computer, it could equally be provided by a Web-based email service.

In this paper we investigate the reliability of multipath routing using multiple gateways in a wireless mesh network. We propose a greedy gateway placement algorithm that tries to minimise the number of gateways needed when a maximum distance between each node and a predefine number of gateways is guaranteed. Simulations results show that our intelligent gateways placement can be employed to complement multipath routing to further improve the reliability of packet delivery.

Routing policies used in the Internet can be restrictive, limiting communication between source-destination pairs to one path, when often better alternatives exist. To avoid route flapping, recovery mechanisms may be dampened, making adaptation slow. Unstructured overlays have been widely proposed to mitigate the issues of path and performance failures in the Internet by routing through an indirect-path via overlay peer(s). Choice of alternate-paths in overlay networks is a challenging issue. Ensuring both availability and performance guarantees on alternate paths requires aggressive monitoring of all overlay paths using active probing; this limits scalability when the number of overlay-paths becomes large. An alternate technique to select an overlay-path is to bias its selection based on physical disjointness criteria to bypass the failure on primary-path. In this paper, we show how type-of-relationship (ToR)-Graphs can be used to select maximally-disjoint overlay-paths.

Routing policies used in the Internet tend to be restrictive, limiting communication between source-destination pairs to one path, when often better alternatives exist. To avoid route flapping, recovery mechanisms may be dampened, making adaptation slow. Unstructured overlays have been widely used to mitigate the issues of path and performance failures in the Internet by routing through alternate paths via overlay peers. The construction of such routing overlays often does not take into account the physical topology of the network, which necessitates that all overlay nodes aggressively probe paths to check performance, limiting scalability. In this paper, we analyze a topologically-aware architecture to estimate end-to-end path availability for service on the Internet. We propose destination-guided detouring via resilient overlay networks (DG-RON); a distributed coordinate-based overlay architecture, which uses landmark based heuristics for scalable end-to-end path discovery. Simulations show that DG- RON can predict alternate paths with a high success rate.

Wireless mesh networks (WMN) have attracted considerable interest in years as a convenient, flexible and low-cost alternative to wired communication infrastructures in many contexts. However, the great majority of research on metropolitan-scale WMN has been centered around maximization of available bandwidth, suitable for non-real-time applications such as Internet access for the general public. On the other hand, the suitability of WMN for mission-critical infrastructure applications remains by and large unknown, as protocols typically employed in WMN are, for the most part, not designed for real-time communications. In this paper, we describe the smart transport and roads communications (STaRComm) project at National ICT Australia (NICTA), which sets a goal of designing a wireless mesh network architecture to solve the communication needs of the traffic control system in Sydney, Australia. This system, known as SCATS (Sydney coordinated adaptive traffic system) and used in over 100 cities around the world, connects a hierarchy of several thousand devices - from individual traffic light controllers to regional computers and the central traffic management centre (TMC) - and places stringent requirements on the reliability and latency of the data exchanges. We discuss our experience in the deployment of an initial testbed consisting of 7 mesh nodes placed at intersections with traffic lights, and share the results and insights learned from our measurements and initial trials in the process.

This paper investigates the problem of selecting multiple routing paths to provide better reliability in multi-radio, multi-channel mesh networks with stationary nodes. Providing reliability is difficult in wireless networks because of unreliable transmissions. Previous work has investigated the use of additional data redundancy to improve the packet delivery rate, but requires node-disjoint paths to be used. In this paper we investigate multipath routing with packet duplication and non-disjoint paths for achieving better reliability and low delay. We propose a path weight function based on the ETT metric and interference minimisation. Our solution provides higher packet delivery ratio and lower end-to-end delay when compared to the single path WCETT metric, a maximally disjoint path selection metric, and the CAM metric.

While email is fairly reliable, messages do get lost. This can be due to the store-and-forward nature of e-mail, which passes responsibility for delivering email from one mail server to another, but in doing so exposes messages to loss, should a server not fulfil its responsibility. Some end users are more sensitive to email unreliability than others, and seek tools with which they can detect and recover from email failures. This paper describes tools that allow email sending software to automatically respond to reports of undeliverability by persisting with retransmission or retransmitting to alternate recipients.

Most research on Distributed Hash Tables (DHTs) assumes ephemeral, lightly loaded deployments. Each node has a lifetime of a few hours and initiates a lookup once every few seconds or minutes. However, in giant internet data centers, each node has a lifetime of weeks or months and initiates hundreds or thousands of lookups every second. In such an environment, one-hop DHTs are superior to multi-hop DHTs. They use lookup bandwidth more efficiently. We qualify conflicting research to show that a single onehop DHT can indeed scale to at least a few hundred thousand nodes in stable, high-capacity enterprise networks. Two new designs are presented: One Hop Sites (1HS), a high-capacity DHT tailored for site redundancy; and the One Hop Federation (1HF), a global, hierarchic DHT that resolves an open latency problem. For both, the analysis a) confirms linear scalability to at least a few hundred thousand nodes and b) identifies the most sensitive design parameters.

This paper describes an algorithm designed to estimate the future cell throughput of a multi-rate wireless network based on measurements. The achievable throughput of a cell depends on the transmission rate used by the mobile nodes. We propose to use a measurement window to gauge the aggregate channel usage by multiple nodes over a period of time instead of predicting the movement and the rate of each node. It is found that the statistics of the average channel occupation time per frame follow the normal distribution. This offers a high degree of confidence in the prediction of cell throughput.

Algorithms by which peers join and leave structured overlay networks can be classified as passive or active. Passive topology maintenance relies on periodic background repair of neighbor pointers. When a node passively leaves the overlay, subsequent lookups may fail silently. Active maintenance has been proven only for fault-free networks. We develop an active topology maintenance algorithm for practical, fault-prone networks. Unlike prior work, it a) maintains ring continuity during normal topology changes and b) guarantees consistency and progress in the presence of faults. The latter property is inherited by novel extension of the Paxos commit algorithm. The topology maintenance algorithm is formally developed using the B method and its event-driven extensions for dynamic systems. Messaging and storage overheads are quantified

The paper investigates fairness in a wireless LAN environment in which stations experience unequal signal qualities due to location and interference factors. This causes differences in transmission rate among mobile stations due to the use of link adaptation schemes. We analyse fairness in the presence of these impairments for the distributed contention based access mechanisms of the IEEE 802.11 and 802.11e standards. The notion of fairness is explored in terms of utility derived from the network (i.e. throughput) and allocated resources (i.e. amount of time permitted to transmit). We show how fairness can be achieved by judicious choice of parameters and study the impact of these choices on capacity. Finally, we demonstrate how these findings could affect other related designs including admission control and charging schemes.

This paper proposes an admission control algorithm that enables the upcoming IEEE 802.11e contention based enhanced distributed channel access (EDCA) to provide quantitative bandwidth guarantees for wireless local area networks (WLANs), rather than a relative prioritized service. The algorithm estimates the throughput that flows would achieve if a new flow with certain parameters was admitted, and so indicates whether such a new flow can be admitted while preserving the quality of service (QoS) of existing flows. The algorithm deals with the EDCA parameters of minimum contention window size and transmission opportunity duration, and indicates what values should be used for different flows. Simulation results confirm the accuracy of the throughput estimates and the effectiveness of the admission control algorithm.

The end-to-end arguments raised by J. Saltzer, D. Reed and D. Clark (see ACM Trans. Comp. Sys., vol.2, no.4, p.277-88, 1984) are amongst the most influential of all communication protocol design guides. However, they have recently been challenged by the advent of firewalls, caches, active networks, NAT (network address translators), multicasting and network QoS. This paper reviews the end-to-end arguments, highlighting their subtleties, and provides additional arguments for and against end-to-end implementations. It shows the importance of trust as a criterion for deciding whether to implement a function locally or end-to-end, and how end-to-end implementations can help robustness, scalability, ease of deployment, and the provision of appropriate services. It focuses on the performance implications of end-to-end or localized functionality and argues against end-to-end congestion control of the form used by TCP.

The IEEE 802.11 MAC protocol supports two modes of operation, a random access mode for non-real-time data applications, and a polling mode for real-time applications. We design and analyze a system that uses the polling mode for interactive voice traffic. With larger inter-poll periods, more voice calls can be accommodated, but at the expense of increased delay. For example, our analysis shows that with an inter-poll period of 90 ms, a maximum of 26 voice calls can be handled with a worst-case delay of 303 ms, whereas with an inter-poll period of 60 ms, a maximum of 17 voice calls can be handled with a worst-case delay of 213 ms. We also carry out an error analysis that demonstrates the need for error correction of voice packets.

In ATM networks, one effective way to provide service guarantees is to shape individual traffic streams at various points in the network. In the paper, the authors present a method for multiplexing a large number of single-stream spacer outputs on cell emissions rather than on cell arrivals. The method not only approximates first come first served (FCFS) multiplexing of such outputs but also better suits bursty arrivals. A number of performance related issues of the proposed method are identified and examined.

Traffic shaping is important in ATM networks, especially those that are interconnected or provide service guarantees. We examine what shaping may be considered ideal, and what is attainable under the constraints of transmission systems and cost. We justify the use of FCFS multiplexing of multiple single-stream shaper outputs as a performance reference for multi-stream shapers, but also point out some of its deficiencies. Shaper implementations in which transmissions are scheduled on cell arrivals, emissions, and transmissions are examined and compared both qualitatively and through simulation. We identify the problem of shaping cells that must conform to multiple traffic constraints (e.g. when the rate of a multicast connection must be adapted to suit multiple links) and examine implementations to achieve this. Shaping in which inevitable cell delay variation is intentionally distributed inequitably amongst connections (to assist CDV-intolerant connections) is also examined.>

Issues arising in the implementation of receivers for asynchronous transfer mode (ATM) networks, e.g., IEEE Std 802.6 MANs (metropolitan area networks) and broadband integrated services digital networks (B-ISDNs), are examined. These include buffering strategies to account for the burstiness and spatial randomness of buffering demands, preventing reassembly deadlock, and enforcing packet reception precedences. It is shown that transnode delays critically affect protocol performance and the complexity of implementing spatial reuse.>