AccessNets

In comparison to existing 3G or 3G+ wireless systems, fourth generation (4G), long-term evolution (LTE) or mobile Wimax are characterized by higher bit rates, highly fluctuant traffic matrices and higher antenna’s density. Current backhauling techniques federating radio antennas are not suited to these new characteristics. Several investigations are carried out for the design of new generation radio access networks (NG-RAN) in charge of concentrating radio cellular traffic from the base stations to the core network. In this paper, we propose an original approach based on an Hybrid Optical Wireless Radio Access Network (HOWRAN) exploiting the benefits of radio-over-fiber technologies and of recent advances in the field of optical devices and systems. As an illustration, we apply the HOWRAN concept to the backhauling of fixed or mobile WiMAX base stations. The two main innovative aspects of HOWRAN are depicted: its hardware architecture and its control plane.

A plethora of enabling optical and wireless technologies have been emerging that can be used to build future-proof bimodal fiber-wireless (FiWi) broadband access networks. After overviewing key enabling radio-over-fiber (RoF) and radio-and-fiber (R&F) technologies and briefly surveying the state of the art of FiWi networks, we introduce an Ethernet-based access-metro FiWi network, called SuperMAN, that integrates next-generation WiFi and WiMAX networks with WDM-enhanced EPON and RPR networks. Throughout the paper we pay close attention to the technical challenges and opportunities of FiWi networks, but also elaborate on their societal benefits and potential to shift the current research focus from optical-wireless networking to the exploitation of personal and in-home computing facilities to create new unforeseen services and applications as we are about to enter the Petabyte age.

Excess bandwidth distribution techniques have recently been proposed to improve the dynamic bandwidth allocation in EPONs. We compare existing offline excess bandwidth distribution with conventional IPACT Limited in terms of packet delay performance. We identify the factors that result in packet delay reduction with excess bandwidth distribution compared to IPACT-Limited and discover that existing offline excess distribution mechanisms become unstable at moderate to high loads in long-range EPONs with large round trip propagation delays. We propose and evaluate a novel Online Excess Bandwidth Distribution (OEBD) mechanism to provide stable excess bandwidth distribution even at high loads in long-range EPONs.

Planning Municipal Wireless Access Networks is a challenging task, since many optimization choices must be taken into account in large metropolitan areas and, in turn, the number of free variables is very large, on the order of several millions, if the position of wireless access points is to be optimized, as well as their connection to the backbone network. This paper formalizes the problem of choosing appropriate access points’ locations and to connect the so built wireless access network to the backbone network. We provide an optimization algorithm able to find a solution with a fast heuristic approach, and apply it to a real-world scenario about a 51 km

2

area of Milano (Italy). The proposed heuristic solution is compared with the result of a simulated-annealing-based optimization algorithm and the result is that the cost of the solutions of the heuristic algorithm is larger than that of the optimal solution by a few percents.

This paper investigates the benefits of Dynamic Spectrum Management (DSM) in terms of reducing the power consumption and improving the data rates in digital-subscriber-line (DSL) networks. The proposed techniques at the three different DSM levels simultaneously also provide a significant improvement in the stability of DSLs. The proposed DSM methods are compared with other non-DSM solutions, which sacrifice power and/or data-rate in order to improve the stability, while also harming other DSLs through impolite power usage. Various examples are presented showing that the proposed DSM methods can avoid such unnecessary impoliteness, and that stability and politeness can be simultaneously achieved in a DSL network that is efficiently managed using DSM.

Orthogonal and quasi-orthogonal codes are integral part of any DS-CDMA based cellular systems. Orthogonal codes are ideal for use in perfectly synchronous scenario like downlink cellular communication. Quasi-orthogonal codes are preferred over orthogonal codes in the uplink communication where perfect synchronization cannot be achieved. In this paper, we attempt to compare orthogonal and quasi-orthogonal codes in presence of timing synchronization error. This will give insight into the synchronization demands in DS-CDMA systems employing the two classes of sequences. The synchronization error considered is smaller than chip duration. Monte-Carlo simulations have been carried out to verify the analytical and numerical results.

We present a simple network design for a large multi-provider municipal WiFi network where the WiFi zones of multiple organizations are merged into a single IP subnet at a central layer 2 switch. The design provides built-in session mobility support for the WiFi clients using the network, without any additional software at the client. We also provide a simple design for mobile access points so that they belong to the same IP subnet. We analyze the session mobility of the over 12000 devices using our network in January 2008. We also provide a comparison of the usage of multi-mode mobile devices and WiFi laptops, and characterize the roaming performance of a standard WiFi laptop in our network.

We propose a modification of the IEEE 802.16e hard handover (HHO) procedure, which significantly reduces the handover latency constraint of the original HHO procedure in IEEE 802.16e networks. It allows a better handling of the delay-sensitive traffic by avoiding unnecessary time-consuming scanning and synchronization activity as well as simplifies the network re-entry procedure. With the help of the backhaul network, it reduces the number of control messages in the original handover policy, making the handover latency acceptable also for real-time streaming traffic. Preliminary performance evaluation studies show that the modified handover procedure is able to reduce the total handover latency by about 50%.

A scheduling algorithm and MAC protocol which provides low-jitter guaranteed-rate (GR) communications between base-stations (BS) in a Wireless Mesh Network (WMN) is proposed. The protocol can provision long-term multimedia services such as VOIP, IPTV, or Video-on-Demand. The time-axis is partitioned into scheduling frames with

F

time-slots each. A directional antennae scheme is used to provide each directed link with a fixed transmission rate. A protocol such as IntServ is used to provision resources along an end-to-end path of BSs for GR sessions. The Guaranteed Rates between the BSs are then specified in a doubly stochastic traffic rate matrix, which is recursively decomposed to yield a low-jitter GR frame transmission schedule. In the resulting schedule, the end-to-end delay and jitter are small and bounded, and the cell loss rate due to primary scheduling conflicts is zero. For dual-channel WMNs, the MAC protocol can achieve 100% utilization, as well as near-minimal queueing delays and near minimal delay jitter. The scheduling time complexity is

O(NFlogNF)

, where

N

is the number of BSs. Extensive simulation results are presented.

In this paper, we investigate coexistence of collocated 802.11 and Bluetooth technologies in 2.4 GHz industrial, scientific, and medical (ISM) band. To that end, we show a time division multiplexing approach suffers from the “avalanche effect”. We then provide remedies to avoid this effect and improve the performance of the overall network. For example, it is shown that a simple request-to-send (RTS) / clear-to-send (CTS) frame handshake in WLAN can avoid “avalanche effect” and improve the performance of overall network.

This paper describes the QoS features that will transform the access networks landscape in order to bring “Broadband” as a commodity while setting up the pillars of the “Future Media Internet”. Quality of Experience is obviously key for emerging and future services. Broadcasting services will first need to equal the QoE of their counterparts in the Open-air market (for IP-TV examples would be artifact-free, no picture freezing, fast zapping times) and offer new features often using interactivity (Time-shifted TV, access to more content, 3DTV with feeling of presence). The huge variety of communications alternatives will lead to different requirements per customer, whose needs will also be dependent on parameters like where the connection is made, the time of the day/day of the week/period of the year or even his/her mood. Today’s networks, designed for providing just Broadband connectivity, will not be enough to satisfy customer’s needs and will necessarily support the introduction of new and innovative services. The Networks of the future should learn from the way the users are communicating, what services they are using, where, when, and how, and adapt accordingly.

An indoor power line is one of the most attractive media for in-home networks. However, there are many technical problems for achieving in-home power line communication (PLC) with high rate and high reliability. One of such problem is the degradation in the performance of the in-home PLC caused by periodically time-varying channel responses, particularly when connecting the switching power supply equipment. We present a measurement method for power line channel responses and reveal the switching of the channel responses synchronized with power-frequency voltage when connecting switching power supply equipment in sending or receiving outlets. In this paper, we term them periodically switching channel responses. The performance of PLC adapters is seriously affected by the periodically switching channel responses. Therefore, we provide a modeling of the periodically switching channel responses by using finite impulse response (FIR) filters with a shared channel memory and construct a simulator for in-home power line channels including the periodically switching channel responses in order to evaluate the various communication systems through the power line. We present the validity of the proposed simulator through the performance evaluation of OFDM/64QAM over periodically switching channels with additive white Gaussian noise. Furthermore, we evaluate the influence of the periodically switching channel responses on the communication quality of a time-invariant modulation scheme by using the proposed simulator.

Broadband Power Line Communications (B-PLC) technology is an alternative for broadband access networks, allowing bit rates up to currently 200Mbps. This technique uses the wiring of the low-voltage grid in order to offer to the users the telecommunications services, such as Internet, VoIP, VoD, etc. The B-PLC design process is sub-divided into two parts: the

Generalized Base Station Placement

(GBSP) problem and the

PLC Channel Allocation Problem

(P-CAP). This paper focuses on GBSP that is modeled as multi-criteria combinatorial optimization problem. Based on our published mathematical modeling, this paper supplies more numerical experiments for the evaluation of Multi-Objective Evolutionary Algorithms (MOEAs) in solving GBSP. Their performance is compared with the single-objective optimization.

The planning process of the Broadband Powerline communications access networks contains two main problem parts: the

Generalized Base Station Placement

(GBSP) problem and the

PLC Channel Allocation Problem

(P-CAP). The GBSP is investigated/solved in our previous works. In this paper, we focus on the P-CAP. The task of the P-CAP consists in allocating a sub-set of channels from an available set of PLC channels to each base station in the B-PLC site. Two optimization objectives are considered for the solution of this problem; namely the maximization of the resource reuse and the minimization of the generated interferences in the site. These objectives are conflicting, since the optimization of one of them results in the deterioration of the other. Therefore, this problem is modeled as a Multi-objective (or multi-criteria) Optimization Problem (MOP). Three variants of Pareto-based multi-objective algorithms, using evolutionary search, are used to solve it. Their performances are evaluated on four problem instances.

As one of the most efficient access solutions, VDSL technology is becoming a highlight in the next generation network. This paper addresses the topology design of hierarchical Hybrid Fiber-VDSL Access Networks (HFVAN) as a NP-hard problem. An efficient strategy with general binary models is proposed to find a cost-effective and high-reliable network with heuristic algorithms in a short time. An enhanced Discrete Binary Particle Swarm Optimization (DBPSO) is developed and successfully implemented for this network planning problem, both for clustering and positioning. In terms of numerical results, the performance of the enhanced DBPSO is compared with some previous approaches.

This work presents a new algorithm for Dynamic Spectrum Management (DSM) in Digital Subscriber Lines. Previous approaches have achieved high performance by attempting to directly solve or approximate the multiuser spectrum optimisation problem. These methods suffer from a high or intractable computational complexity for even a moderate number of DSL lines. A new method is proposed which is a heuristic extension of the single user greedy algorithm applied to the multi-user case. The new algorithm incorporates a novel cost function that penalises crosstalk as well as considering the

usefulness

of a tone. Previous work has proved the performance of the new algorithm in simple 2-user scenarios. In this work we present new results which demonstrate the performance of the algorithm in larger DSL bundles. Simulation results are presented and it is shown that the new method achieves results within a few percent of the optimal solution for these scenarios.

The paper discusses the merits of having a reconfigurable access network. The network is viewed as a stack of logical PONs in which bandwidth can be redistributed on an inter-PON scale. The redistribution allows for optimal distribution of bandwidth to the end user.

Gigabit passive optical network (GPON) is likely to play an important role in future access networks and the current challenge is to increase the existing GPON bit-rate to 10 Gb/s to provide next generation access (NGA). However, implementing this in a cost-effective manner is difficult and an important research topic. One of the difficulties in implementation for the electronic part of high-speed GPON is the fragmentation feature as it requires multiple pipeline paths. This paper proposes a novel segmentation and reassembly (SAR) scheme, which simplifies the implementation of fragmentation in that it employs fewer FPGA resources and allows a faster hardware clock rate. Analysis confirms that the scheme does not suffer from reduced efficiency in a variety of conditions. It is also backward compatible and suitable for current 1.25 Gb/s and 2.5 Gb/s GPONs. The novel SAR is verified by both a hardware GPON emulator and a software OPNET simulation.

The deployment of optical access networks is considered by many as the sole solution able to cope with the ever-increasing bandwidth needs of data and media applications. Gigabit-capable Passive Optical Networks (GPON) are being adopted by many operators worldwide as their preferred fiber-to-the-home network architecture. In such systems, the Medium Access Control (MAC) layer is a key aspect of their operation and performance.

TCP is the transport protocol of choice of most popular applications. However, TCP performance is known to be sensitive to the behavior of MAC-layer mechanisms. Thus, it is important to assess the impact that the GPON MAC layer may have on TCP. Motivated by this, in this paper we present a preliminary study of TCP performance issues that may arise in GPON networks. Based on a simple system model, the interaction of some GPON MAC features with TCP is explored both analytically and by simulation.

This paper presents a fast channel switching method in Ethernet Passive Optical Network (EPON) system for IPTV service. Fast channel switching is one of the important features of successful IPTV systems. Users surely prefer IPTV systems with small channel switching time rather than a longer one. Thus a channel switching control module and a channel/permission list in EPON system’s ONU or OLT is designed. When EPON system receives channel switching message from IPTV end user, the channel switching control module will catch the message and search the channel list and permission list maintained in EPON system, then got the matching parameter of EPON for the new channel. The new channel’s data transmission will be enabled by directly updating the optical filter of the ONU that end user connected. By using this method in EPON system, it provides a solution for dealing with channel switching delays in IPTV service.

WDM based dynamic bandwidth allocation schemes are proposed as potential future-proof upgrade solutions for access PON systems. The schemes represent cost-efficient ways to exploit the virtually unlimited bandwidth afforded by passive optical architectures whilst remaining fully compatible with the IEEE Ethernet standard for the first mile (EFM 802.3ah).

WiMAX system is based on OFDMA, and it is very suitable for VoIP traffic. At the same time, aggressive AMC and frame bundling are two important techniques to improve the performance of VoIP, and so they are applied to WiMAX system in this paper. The simulation results show that when the system is in short of bandwidth, these two techniques indeed reduce the PER and delay jitter of the VoIP users, and also enhance the spectrum efficiency.

Multi-user orthogonal frequency division multiplexing (OFDM) with adaptive sub-carrier and bit allocation are considered in this paper. Assuming having the knowledge of the instantaneous users’ channel gains, three typical dynamic sub-carrier and bit allocation algorithms are analyzed and compared. The goal of each algorithm is to minimize the total transmit power with the user’s data rate constraint. Then, based on the comparison, an improvement method is proposed for Zhang algorithm. Finally, performance comparisons between adaptive schemes and static ones are given. The results prove that the performances of adaptive algorithms are much better than that of the traditional fixed sub-carrier allocation method.

In this paper, we propose a new wireless scheduling algorithm for the IEEE 802.11e HCF. The algorithm grants the mobile stations variable time for the upstream data flow in proportion to the queue size of the transmission buffer. At the same time, it retrieves half of the extra time allocated in previous cycle from those flows whose requirement has stopped. Hence, the system achieves stability by preventing the ping-pong phenomena. The algorithm is computationally simple, and as compared to the other algorithms, it gives bounded delays and jitter to the real-time applications under heavy load conditions.

Choosing the best access technology for delivering broadband services, is an important and complex step in moving toward NGN, specially in non-developed and developing countries. This paper, proposes a simple classification of users due to their communicational needs, and recommends the most appropriate access method for each class of users, considering network infrastructure in non-developed and developing countries, and capabilities and shortcomings of current technologies.

In this paper, we provide an analysis to the performance of optical time-wavelength code-division multiple-access (OTW-CDMA) network when the system is working above the nominal transmission rate limit imposed by the passive encoding-decoding operation. We address the problem of overlapping in such a system and how it can directly affect the bit error rate (BER). A unified mathematical framework is presented under the assumption of one coincidence sequences with non-repeating wavelengths. A closed form expression of the multiple access interference limited BER is provided as a function of different system parameters. Results show that the performance of OTW-CDMA system may be critically affected when working above the nominal limit; an event that may happen when the network operates at high transmission rate. In addition, the impact of the derived error probability on the performance of two newly proposed MAC protocols, the S-ALOHA and the R3T, is also investigated. It is shown that for low transmission rates, the S-ALOHA is better than the R3T; while the R3T is better at very high transmission rates. However, in general it is postulated that the R3T protocol suffers a higher delay mainly because of the presence of additional modes.

In this paper, we experimentally study the effects of using low cost optical amplifiers and lasers in multi-format multi-wavelength radio over fiber signals. We analyze the propagation of UMTS, WLAN and WiMAX radio signals in a single channel scenario and study the impact of amplifying the referred radio signals together with amplitude modulated ones, which can be ethernet, in WDM scenario.

The latest digital subscriber line (DSL) technology, VDSL2, used for broadband access over twisted-pairs, promises up to 100 Mbit/s for both transmission directions on short loops. Since these systems are designed to operate in a far-end crosstalk (FEXT) limited environment, there is a severe performance degradation when deployed in distributed network scenarios. With power back-off (PBO) the network operators attempt to protect modems deployed on long loops by reducing the transmit power of the short ones. However, currently very little guidance has been given to operators on how to set and optimize the parameters for PBO. In this paper we explore one promising method, the cable bundle unique PBO (CUPBO), which optimizes these parameters according to the actual situation in the cable with regard to noise and network topology. Using real VDSL systems and cables we show that CUPBO algorithm achieves a significant increase in performance compared to the case when one naively takes the PBO values given in the VDSL standard.

QoS Provisioning for 3P over xDSL remains a challenging task due to the effects of line impairments on such services. Differently from simple data, video and voice services have strict requirements for loss and delay tolerance. The accurate assessment of final service quality is part of this provisioning process, but its direct measurement is yet not practical. In this paper we explore the possibility of estimating service quality, with focus on video delivery, by investigating its relationship with performance data available to xDSL operators and deriving models for estimating quality from this data. Experiments using a real xDSL platform and different noise types were conducted. The derived models showed to be accurate enough to estimate video quality for the scenarios evaluated.

Recent research activities about hybrid optical wireless networks become attractive. In this paper, we provide an overview on integration of an Ethernet passive optical network (EPON) and a worldwide interoperability for microwave access (WiMAX) network. Gateway node that integrates both optical scheduling and wireless scheduling functions gains most interests. We promote a question: how the scheduler at optical network unit is affected by its wireless connections when ONUs transfer packets to optical line terminal. Scheduling algorithms have been proposed intensively in EPON system. However, the existing scheduling algorithms are only for either pure optical networks or pure wireless networks. Within this paper, we propose to design packet scheduling schemes that take wireless channel condition into account for the hybrid optical wireless networks. Our scheduling scheme is based on a cost function, which considers not only the packet transmission but also the radio link condition.

Considering the number of access network technologies and the investment needed for the “last mile” of a solution, in today’s highly competitive markets, planning tools are crucial for the service providers to optimize the network costs and accelerate the planning process. In this paper, we propose to tackle the problem of planning access networks composed of four technologies/architectures: the digital subscriber line (xDSL) technologies deployed directly from the central office (CO), the fiber-to-the-node (FTTN), the fiber-to-the-micro-node (FTTn) and the fiber-to-the-premises (FTTP). A mathematical programming model is proposed for this planning problem that is solved using a commercial implementation of the branch-and-bound algorithm. Next, a detailed access network planning example is presented followed by a systematic set of experiments designed to assess the performance of the proposed approach.

Digital subscriber lines offer the possibility to deliver broadband services over the existing telephone network. Still, beforehand subscriber loops must be tested to see whether they can support high-speed data services, and at what bit rate. From the existing measurement techniques, Single Ended Line Testing is often preferred because all necessary measurements can be performed from the central office. Consequently the capacity cannot be measured directly, but should be calculated through the estimation of the loop make-up. This paper discusses some main difficulties of this identification. Moreover, in contrast to the traditional approach where the data are interpreted in the time domain, this paper presents a new approach by doing most of the processing in the frequency domain.

A spectral-polarization coding (SPC) optical code-division multiple-access (OCDMA) configuration structured over arrayed-waveguide grating (AWG) router is proposed. The polarization-division double balanced detector is adopted to execute difference detection and enhances system performance. The signal-to-noise ratio (SNR) is derived by taking the effect of PIIN into account. The result indicates that there would be up to 9-dB SNR improvement than the conventional spectral-amplitude coding (SAC) structures with Walsh-Hadamard codes. Mathematical deriving results of the SNR demonstrate the system embedded with the orthogonal state of polarization (SOP) will suppress effectively phase-induced intensity noise (PIIN). In addition, we will analyze the relations about bit error rate (BER) vs. the number of active users under the different encoding schemes and compare them with our proposed scheme. The BER vs. the effective power under the different encoding scheme with the same number of simultaneous active user conditions are also revealed. Finally, the polarization-matched factor and the difference between simulated and experimental values are discussed.

As large-scale wireless networks continue to proliferate, a reliable way to test coverage and communicate requirements becomes increasingly important. In this paper we discuss concerns and provide guidelines to consider when developing a coverage testing methodology for large-scale wireless networks. We propose a method which complies with these guidelines and apply it to a large municipal mesh network in Portland, Oregon. This approach is at the same time simple, cost-effective, and rigorous. We use commodity hardware to perform both high and low-layer tests at a random sample of points. Our method provides insights into the “expected” performance and coverage of the network. We find that a greater density of nodes is required in Portland to provide the required level of coverage, but that at coverered areas, the performance is within specification. We are able to make these extrapolations with high statistical confidence, on a large network, using only 53 measurement points and a single measurement device which cost less than $200 USD to build.