Accurately assessing the number of people in a room is essential for enhancing operational efficiency, safety, and sustainability, enabling applications such as smart building management, energy conservation, and emergency evacuation planning. Wi-Fi sensing, favored for its privacy-preserving capabilities and the ubiquity of Wi-Fi infrastructure, has become a popular method for such sensing tasks, including people counting. Within this context, our paper introduces a Convolutional Neural Network (CNN) model for millimeter wave (mmWave) Wi-Fi people counting, capitalizing on the IEEE 802.11bf amendment, and in particular its passive sensing framework. By evaluating Range-Doppler (RD), Azimuth-Doppler (AD), and AzimuthRange (AR) map representations as inputs, we establish AR maps’ superiority, achieving up to $\mathbf{9 8 . 5 7 \%}$ accuracy for counting up to four individuals. Our solution, free of communication overhead, also minimizes energy consumption and computational requirements, offering a scalable and efficient sensing solution for people counting, which is particularly well-suited for Internet of Things (IoT) devices with limited resources.

Wi-Fi sensing has been used to detect and track movements in an environment, resulting in the emergence of several innovative applications. Wi-Fi sensing can detect movement and locate objects by analyzing variations in the Wi-Fi signal due to its interaction with moving objects. Until recently, Wi-Fi sensing has been primarily available through proprietary solutions, which has limited its adoption. However, the recent initiative by the IEEE to develop the IEEE 802.11bf standard promises to make the adoption of Wi-Fi sensing widespread. Although Wi-Fi sensing procedures in communication standards can be overhead, there is currently a lack of literature exploring the sensing performance of Wi-Fi sensing procedures specified in the IEEE 802.11bf standard and its impact on data communication. Therefore, this paper presents a comprehensive evaluation of the sensing performance of the IEEE 802.11bf protocol and its impact on data communication in different configurations. Our findings expose the limitations of specific configurations and pave the way to provide guidance on efficient operating configurations of an IEEE 802.11bf network.

The next generation WiFi such as IEEE 802.11ad and 802.11ay can provide stringent Quality of Service (QoS) due to its support of contention free channel access called Service Period. IEEE 802.11ad supports two types of user traffic: isochronous and asynchronous. These user traffic need guaranteed Service Period duration before their periods. Hence, admission control plays an important role in an IEEE 802.11ad system. In an earlier work we studied admission control only for isochronous requests. In this paper, we present admission control and scheduling algorithms which can handle both types of requests. We devise a proportional fair and linear run time complexity algorithm that treats asynchronous requests as periodic requests, because of which it overallocates resources to the asynchronous requests. The conditions of possible performance loss due to this overallocation are analyzed. We provide detailed simulation results which show that presence of asynchronous request degrades performance of isochronous requests in terms of number of admitted requests and channel utilization. But, the smaller number of admitted isochronous requests perform better in terms of channel allocation time and delay.

An upsurge of low latency and bandwidth hungry applications such as virtual reality, augmented reality and availability of unlicensed spectrum in the millimeter wave band at 60 GHz have led to standardization of the new generation WiFi systems such as IEEE 802.11ad and 802.11ay. Due to the stringent Quality of Service requirement of those applications, IEEE 802.11ad/ay have introduced contention free channel access calledService Period. One type of user traffic supported by IEEE 802.11ad is isochronous traffic, which is essentially periodic traffic that requires certain channel time to be allocated before its period ends. In an earlier work, we presented three Admission Control Algorithms (ACAs) which admit isochronous requests to achieve the above goals. One of these ACAs, the proportional fair allocation admission control (PFAAC), offers the best tradeoff across different performance metrics. But it did not consider guard time (GT) overhead, which is essential in a practical system. In this paper, we present two methods to compute upper bounds on GT overhead. We evaluate performance of the modified PFAAC with the two methods and PFAAC with no GT overhead. The modified PFAAC with the method that uses a tighter upper bound on GT overhead, provides the best performance.

The ubiquitous deployment and availability of wireless communications devices, coupled with recent technical advancements, provide a unique opportunity to enable wireless sensing applications, leveraging existing communications equipment and signals. The availability of modeling tools and dataset is crucial to support the development of sensing techniques and to understand the end-to-end performance of a joint wireless communication and sensing system. However, most of the sensing performance evaluations are carried out using proprietary tools and dataset. In this paper, we present a set of open source tools and models enabling the evaluation of future WLAN sensing systems. Our framework is composed of a ray-tracing implementation specific for sensing application, an IEEE 802.11ay physical layer (PHY) digital transceiver model and a visualization application. Using these tools, we design a dataset consisting of more than 14 000 entries of millimeter wave channels and IEEE 802.11 ay signals to democratize the design of both data-driven and model driven communication and sensing algorithms. We also provide a preliminary evaluation of a CSI-based WLAN sensing system using IEEE 802.11 ay signals. The results indicate that existing communication systems can be used to enable sensing applications.

An upsurge of low latency and bandwidth hungry applications such as virtual reality, augmented reality and availability of unlicensed spectrum in the mmWave band at 60 GHz have led to standardization of the next generation WiFi such as IEEE 802.11ad and 802.11ay. Due to the stringent Quality of Service (QoS) requirement of those applications, 802.11ad/ay have introduced contention free channel access called Service Period, which provides dedicated channel access exclusively reserved for communication between a pair of nodes. One type of user traffic supported by IEEE 802.11ad is isochronous traffic, which is essentially periodic traffic that requires certain channel time to be allocated before its period ends. So, isochronous traffic needs guaranteed channel time allocation with stringent deadlines. In this paper, we present three Admission Control Algorithms (ACAs) which admit isochronous requests to achieve the above goals while being fair. We also present an Earliest Deadline First (EDF) based scheduling algorithm for isochronous traffic. We evaluate the performance of the three ACAs in terms of different performance metrics. Our simulation results show that, out of the three ACAs, the proportional fair allocation based algorithm offers the best tradeoff across different performance metrics.

In this paper, we present a machine learning (ML) based dynamic spectrum access (DSA) scheme which can be used in a system in which the primary user (PU) spectrum occupancy can be represented as a sequence of busy (on) and idle (off) periods. We use real world data collected from Long Term Evolution (LTE) systems at two locations for our study. We experiment with different feed forward artificial neural network (ANN) architectures to choose from for our DSA scheme. A simple perceptron based ANN architecture was determined to provide good performance. We compare performance of our ML based DSA scheme with a traditional DSA scheme based on analytical model that uses survival analysis. Our results show that our ML based scheme outperforms the survival analysis based scheme in terms of utilization of idle periods. In terms of probability of interference to the PU, our scheme is better in some configurations and slightly worse in some other configurations.

Dynamic Spectrum Access (DSA) promises to be a shared spectrum technology that can alleviate the artificial spectrum crunch created by the static allocation of spectrum. There have been many studies on DSA systems in the literature. However, most of them are analyzed using simulation studies rather than analytical modeling. Analytical models are useful in evaluating performance of such systems quickly and easily. In this paper, we present an analytical model of an opportunistic DSA system. Using an alternating renewal process to represent primary user traffic of the DSA system and applying theory from survival analysis and stochastic process, we derive an expression to compute the white space utilization (WSU) of a DSA system for a general distribution of idle durations of primary traffic. Taking the exponential distribution as an example, we validate our analytical model by comparing its results to results obtained from two simulation experiments. One experiment uses idle durations generated from pseudorandom variates and the other uses data collected from a real Long-Term Evolution (LTE) system whose idle duration distribution is approximately exponential. Our analytical WSU results match closely with those from the first experiment and match reasonably well with those obtained from the second experiment.

Commercial-federal spectrum sharing in the 3.5 GHz band requires an Environmental Sensing Capability (ESC) system, consisting of sensors deployed along the coasts, to detect federal incumbent shipborne radar in order to protect it from harmful interference from commercial users. The coastal waters where the radar needs protection from interference are divided into a chain of contiguous polygons called Dynamic Protection Areas (DPAs). The sensor(s) associated with each DPA must cover that DPA completely but should minimize any excess coverage on land, in neighboring DPAs, and out at sea. Thus, placement of sensors and their operating parameters are determined by solving this coverage problem. We use existing tower sites as candidate locations for the sensors and the Irregular Terrain Model (ITM) in point-to-point mode to compute the path loss. We present an algorithm for computing the locations and operating parameters of the sensors such that the excess area, and thus the probability of false alarm, is minimized.

In this study, we design and implement an algorithm for optimal dynamic spectrum access (DSA) in a shared spectrum system where the primary user (PU) is a Long Term Evolution (LTE) system. The cumulative hazard function from survival analysis is used to predict the remaining idle time available in each channel for secondary user (SU) transmission subject to a probability of successful completion. Optimal allocation of physical resource blocks (PRBs) for the SU is shown to be a variation of the unbounded knapsack problem. We evaluate the algorithm performance using three data sets collected from real LTE systems. The algorithms achieve good white space utilization and have a measured probability of interference around the target threshold.

The Federal Communications Commission rules for operation in the 3.5 GHz band require that an Environmental Sensing Capability (ESC) system detect the presence of a federal incumbent shipborne radar in order to protect it from harmful interference. Thus, ESC operators have to deploy ESC sensors along the coasts to comply with the rules. We formulate the ESC sensor deployment problem as a coverage problem where ESC sensors need to cover a predefined geometric area inside which radar may experience harmful interference. Using propagation models and radar parameters, we compute antenna lobe patterns for different beamwidths and detection thresholds of the ESC sensors. These patterns are then used to cover the geometric area such that both outage and excess coverage areas are minimized. We present a greedy algorithm and apply it to Dynamic Protection Areas currently being defined for the coasts of the contiguous United States. We evaluate its performance in terms of some key metrics important to the federal incumbent as well as commercial operators.

In this study, we design and implement two algorithms for dynamic spectrum access (DSA) that are based on survival analysis. They use a non-parametric estimate of the cumulative hazard function to predict the remaining idle time available for secondary transmission subject to the constraint of a preset probability of successful completion. To show that the algorithms are effective in real-world scenarios even at fine time scales, we evaluate their performance using data collected from an LTE band to model primary user activity. The algorithms are run in different configurations, i.e., they are trained and run on a few combinations of data sets. Our results show that as long as the cumulative hazard functions are fairly similar across datasets, the algorithms can be trained on one day's dataset and run on that of another day's without any significant degradation of performance. The algorithms achieve fairly high white space utilization and have a measured probability of interference which always stays below the preset threshold.

Traffic engineering (TE) has long been used by network providers to reduce network congestion and improve resource utilization. Due to its significance, several traffic engineering algorithms have been proposed in literature. However, most of these algorithms optimize maximum link utilization (MLU) in network, and/or assume that network has the capability to route demands on arbitrary paths. Optimizing only for MLU can result in longer route computations to save bandwidth along shorter paths, thereby hurting application performance (as shown by recent research). Further, minimizing MLU can lead to solutions where several links have utilization close to MLU, while many others are under- utilized. Besides, as large fraction of today's Internet uses OSPF routing protocol, it cannot benefit from TE algorithms assuming arbitrary routing capabilities. To address these problems, we present Wise-OSPF (WOSPF), a traffic engineering solution for OSPF networks. WOSPF formulates TE as an optimization problem. The objective of WOSPF is to minimize the difference between the maximum and minimum link utilizations across the network, which leads to more uniform traffic distribution compared to optimizing MLU. As WOSPF uses OSPF for routing demands, it does not compute unnecessarily long routes and can be employed in legacy OSPF networks with minimal changes. Our results show that WOSPF reduces standard deviation of link utilizations in network by 31.35% compared to an optimal MLU based TE approach, while achieving an MLU within 1.9% of the optimal.

Opportunistic Spectrum Access (OSA) is one of the models proposed in the literature for Dynamic Spectrum Access (DSA). Providing disruption QoS in terms of interference caused to the Primary Users (PUs) is crucial in such systems. In this paper, we use a residual idle time based scheme called RIBS, to provide disruption QoS to the PUs. The transmission duration of a Secondary User (SU) is carefully computed such that the interference to the PU remains below a threshold. We propose two disruption Quality of Service (QoS) metrics and show the computation of maximum transmission duration for the two metrics. We have implemented RIBS with these two new disruption QoS metrics using GNU Radio over Universal Software Radio Peripheral (USRP) hardware. We also use maximum likelihood estimation (MLE) to dynamically estimate the parameter of PU traffic, which is then used to compute SU transmission duration. This eliminates the requirement, widely assumed in the literature, that SUs know the PU traffic characteristics a priori. Results from our experiments show that RIBS is able to provide the required QoS and that the MLE based parameter estimation method works reasonably well.

We present a stochastic model-based opportunistic channel access and transmission scheme for cognitive radio enabled secondary users for single data channel. We refer to this scheme as Residual Idle Time Distribution based Scheme (RIBS). In this scheme, the SU randomly senses the channel and if the channel is sensed idle, then it uses residual idle time distribution to estimate its transmission duration such that the probability of its interference with PU is below a predefined threshold. We derive analytical formulae for various performance metrics of SU and validate them through simulations. Simulation experiments are conducted for two different scenarios. In the first scenario, we use synthetically generated channel occupancy data due to PU transmissions using two standard distributions (2-phase Erlang and Uniform distribution). In the second scenario, we use data collected by simulating a TDMA-based PU network that runs realistic applications (VoIP and Web browsing). A pair of sender and receiver SU uses RIBS to opportunistically transmit on the channel. Our simulation experiments show that SU can use RIBS to opportunistically transmit on the channel without violating the interference probability constraint. We also list some of the challenges in using RIBS in realistic scenarios, and provide a comprehensive methodology to use RIBS in such scenarios.

In this paper, we discuss the problem of non-collocated coexistence of WiFi and 4G technologies such as WiMAX and LTE due to adjacent channel interference. The existing literature has many solutions and schemes to address the problem of shared channel coexistence and adjacent channel coexistence on multi-radio platforms. Results for Non-collocated coexistence in adjacent channels in wireless remain very scattered and few. Radio devices operating on Broadband Wireless Access (BWA) 4G wireless technologies like IEEE 802.16 (WiMAX) and LTE-A require very low noise floor. BWA spectrum allocations in 2.3 GHz and 2.5GHz have resulted in these networks to be very close to 2.4 GHz ISM band used by WiFi. We show, with measurements on our test-bed and from existing results, that the low-cost filters on WiFi devices are not very effective in controlling the out-of-band emissions to satisfy the low noise floor requirements of 4G. We propose schemes to mitigate the problem of adjacent channel interference by a time sharing mechanism across technologies by protecting packet receptions on both IEEE 802.11 and the IEEE 802.16 side. We demonstrate the effectiveness of our scheme to protect WiMAX packets by ensuring a controlled silence zone in the WiFi network using a test-bed. We also show that there is very limited adverse impact, due to the use of our scheme, on the system throughput of the non-collocated WiFi network operating in the adjacent channel.

In this paper, we present the methodologies used in existing literature for Virtual Machine (VM) placement, load balancing and server consolidation in a data center environment. While the methodologies may seem fine on the surface, certain drawbacks and anomalies can be uncovered when they are analyzed deeper. We point out those anomalies and draw backs in the existing literature and explain what are the root causes of such anomalies. Then we propose a novel methodology based on vector arithmetic which not only addresses those anomalies but also leads to some interesting theories and algorithms to tackle the above mentioned three functionalities required in managing resources of data centers. We believe that with a strong mathematical base, our methodology has the potential to become the foundation of future models and algorithms in this research area.

There has been a considerable body of research devoted to the design and performance of PC based software routers running open source software. Most of the research on open software routers (OSRs) have focused on improving the performance of single PCs with a few proposals for a distributed design. Modern routers are equipped with enhanced functionality such as QoS features in addition to packet forwarding. However, providing enhanced functionality in distributed OSR architectures has largely remained unaddressed. A distributed design introduces challenges in its implementation due to looser coupling between different subsystems. WFQ is a widely used scheduler that enables QoS features in a router. The inherent centralized nature of the design of WFQ schedulers in most switches and routers creates several challenges when exported to distributed architectures. In this paper, we study the challenges of implementing WFQ in a distributed OSR, propose some novel techniques to address these challenges and compare the performance of our WFQ implementation in distributed OSR with that of a centralized WFQ scheme.

Flooding a network with a message from the sink is required for many purposes like synchronization, code dissemination etc. While several flooding schemes exist, only a few are designed to achieve the energy efficiency required by Wireless Sensor Networks (WSNs). In this paper, we present a TDMA-based Reliable and Energy Efficient Flooding Protocol (TREEFP) for WSNs. Slot assignment in TREEFP is done such that the time taken to flood the network is bounded to a single TDMA frame. TREEFP has a tunable system parameter which brings in tradeoff between reliability, flooding delay and energy consumption because when this parameter changes, the topology of the logical flooding tree also changes. We provide details of simulation experiments to compare TREEFP with other flooding protocols in the literature like FTSP, TDFS and MST. Simulation results show that TREEFP is better than FTSP and TDFS in terms of energy and flooding delay and comparable to MST in terms of those metrics. In terms of reliability, TREEFP is better than MST and comparable to FTSP.

We propose an opportunistic channel access scheme for cognitive radio-enabled secondary networks. In our work, we model the channel occupancy due to Primary User (PU) activity as a 2-state Alternating Renewal Process, with alternating busy and idle periods. Once a Secondary Node (SN) senses the channel idle, the proposed scheme uses the residual idle time distribution to estimate the transmission duration in the remaining idle time, subject to an acceptable PU interference constraint. The SN transmits the frames within the transmission duration without further sensing the channel, thereby reducing the average sensing overhead per transmitted frame. The analytical formulation used by the scheme does not require the SN to keep track of the start of the idle period. We validate the analytical formulations using simulations, and compare the performance of the proposed scheme with a Listen-Before-Talk (LBT) scheme.

We describe an opportunistic channel access scheme for cognitive radio-enabled secondary nodes (SNs). The proposed scheme uses the residual channel idle time distribution to estimate the transmission duration in the remaining idle time, subject to an acceptable Primary User (PU) interference constraint. The SN then transmits the frames within the estimated duration without further sensing the channel, which reduces sensing overhead. The scheme does not require the SN to continuously sense the channel to keep track of the start of the idle period, thereby conserving energy.

This paper presents an integrated MAC and routing protocol called Delay Guaranteed Routing and MAC (DGRAM) for delay-sensitive wireless sensor network (WSN) applications. DGRAM is a TDMA-based protocol designed to provide deterministic delay guarantee in an energy-efficient manner. The design is based on slot reuse to reduce latency of a node in accessing the medium, while ensuring that the medium access is contention-free. The transmission and reception slots of nodes are carefully computed so that data is transported from the source toward the sink while the nodes could sleep at the other times to conserve energy. Thus, routes of data packets are integrated into DGRAM, i.e., there is no need for a separate routing protocol in a DGRAM network. We provide a detailed design of time slot assignment and delay analysis of the protocol. We have simulated DGRAM using ns2 simulator and compared the results with those of FlexiTP, which is another TDMA protocol that claims to provide delay guarantee, and with those of a basic TDMA MAC. Simulation results show that the delay experienced by data packets is always less than the analytical delay bound for which the protocol is designed. Also, the TDMA frame size with DGRAM is always lesser compared to that of FlexiTP, which makes the maximum possible delay much lesser than that of FlexiTP. The average delay experienced by packets and the average total energy spent in the network are much lesser in a network using DGRAM than that using FlexiTP or the basic TDMA MAC.

We propose a model-based channel access mechanism for cognitive radio-enabled secondary network, which opportunistically uses the channel of an unslotted primary network when the channel is sensed idle. We have considered IEEE 802.11 WLAN as a de facto primary network operating in ISM band. Our study focuses on a single WLAN channel that is used by WLAN clients and a WLAN server for a mix of Email, FTP, and HTTP-based Web browsing applications. We model the occupancy of the channel by primary WLAN nodes as an alternating renewal process. The secondary node uses the model to estimate residual idle time duration after the channel is sensed idle, and opportunistically transmits frames in that duration. Our simulation results show that the performance of secondary network is sensitive to the channel sensing duration and that high secondary throughput can be achieved without affecting the primary network significantly by choosing appropriate value of channel sensing duration.

In this paper, we consider a non-saturated IEEE 802.11 based wireless network. We use a three-way fixed point to model the node behavior with Bernoulli packet arrivals and determine closed form expressions for the distribution of the time spent between two successful transmissions in an isolated network. The results of the analysis have been verified using extensive simulations in QualNet. The methodology presented in the paper is novel and we believe that the analysis like ours can be used as an approximation to model the behavior of sub-components of a larger mesh or hybrid network.

In a wireless sensor network (WSN), interesting events are reported to the sink by the sensors in a distributed manner. Applications running at the sink require certain reliability in terms of events per unit time to be able to run satisfactorily. Individual report from sensor nodes is not important, but collective reports from sensor nodes of a region of interest of an application are crucial. Thus, the application event rate can be split across different sensor nodes so as to optimize the usage of scarce resources in WSN such as battery power and memory. In this paper, we propose a transport protocol which provides the desired event reliability to the application, by distributing the load at a sensor among its children based on their residual energies and average MAC layer data rate. The event rate distribution happens in such a way that the application at the sink gets its required event rate and the overall energy consumption of nodes is minimized. This protocol can be used for any MAC protocol as long as the average MAC data rate is known. We take the example of two MAC protocols, Slotted CSMA and Probabilistic TDMA. We derive a method for computing average MAC data rate for these two protocols and then show, using simulations, that our transport protocol performs close to optimal.

This paper presents an integrated MAC and routing protocol called Delay Guaranteed Routing and MAC (DGRAM) for delay sensitive wireless sensor network (WSN) applications. DGRAM is a TDMA-based protocol designed to provide deterministic delay guarantee in an energy efficient manner. The design is based on slot reuse to reduce the latency of a node in accessing the medium, while ensuring contention free medium access. The transmission and reception cycles of nodes are carefully computed so that data is transported from the source towards the sink while the nodes could sleep at the other times to conserve energy. Thus, routes of data packets are integrated into DGRAM. We provide a detailed design of time slot assignment and delay analysis of the protocol. One major advantage of DGRAM over other TDMA protocols is that the slot assignment is done in a fully distributed manner making the DGRAM network self-configuring. We have simulated DGRAM using ns2 simulator and compared the results with those of SMAC for a similar network. Simulation results show that the delay experienced by data packets is always less than the analytical delay bound for which the protocol is designed. As per simulation results, the average energy consumption does not change as the event rate changes, and is less than that of SMAC. This characteristic of DGRAM provides flexibility in choosing various operating parameters without having to worry about energy efficiency.

Open Shortest Path First (OSPF) is one of the most widely used intra-domain routing protocol. It is well known that OSPF protocol does not provide flexibility in terms of packet forwarding to achieve any network optimization objective. Because of the high cost of network assets and commercial and competitive nature of Internet service provisioning, service providers are interested in performance optimization of their networks. This helps in reducing congestion hotspots and improving resource utilization across the network, which, in turn, results in an increased revenue collection. One way of achieving this is through Traffic Engineering. Currently traffic engineering is mostly done by using MPLS. But legacy networks running OSPF would need to be upgraded to MPLS. To achieve better resource utilization without upgrading OSPF network to MPLS is a challenge. In this paper we present a simple but effective algorithm, called Smart OSPF (S-OSPF) to provide traffic engineering solution in an OSPF based best effort network. We formulate an optimization problem based on the traffic demand to minimize the maximum link utilization in the network. Routing of the traffic demand is achieved using OSPF. We have simulated S-OSPF on real networks of two service providers. Simulation results show that S- OSPF based traffic engineering solution performance very closely follows the optimal solution.

Scheduling and call admission control (CAC) in IEEE 802.16 system play a vital role in the performance of the system. The 802.16 standard does not specify any scheduling architecture or CAC. Many proposals assume a bandwidth based CAC which only provides bandwidth guarantee, but cannot fulfill delay and jitter requirements. In this paper, we propose a CAC that ensures QoS guarantee in terms of bandwidth, delay and jitter. We also present a novel method of estimating bandwidth requirement of variable bit rate application to increase the resource utilization of the system. We present our simulation result to show the effectiveness of bandwidth estimation and better performance over bandwidth based CAC.

This paper presents RTMAC, a realtime MAC protocol for wireless sensor network that can provide delay guarantee. RTMAC is based on TDMA protocol, but it is carefully designed to overcome the high latency of traditional TDMA protocols. It also conserves energy when a node may not be transmitting or receiving packets. We discuss the details of time slot assignment procedure of RTMAC and then present delay analysis of the protocol. We compare the performance of RTMAC with the well known energy efficient MAC protocol S-MAC using simulation. The simulation results show that RTMAC is better than S-MAC in terms of providing delay guarantee to packets.

Telephony gateways are devices that interface between IP telephony networks and the PSTN. A telephony proxy or location server (LS) that attempts to connect a call for an end user may need to choose between multiple candidate gateways. Gateways need to send updates of their dynamic resource information periodically to the LS which is used by LS while choosing a gateway. Telephony gateway registration protocol (TGREP) is an IETF proposed standard used by gateways for this purpose. To the best of our knowledge, there is no study done on the performance of TGREP and we believe that this is the first paper on performance modeling of TGREP system. In this paper, we propose an analytical model for TGREP system and then use the model to evaluate different gateway selection algorithms proposed by us in this paper. We validate our model using simulation results. We also present an adaptive update mechanism in which the overhead of sending dynamic resource information is reduced without significant effect on call blocking probability. We outline a few important conclusions from this study that are quite contrary to what IP telephony providers might resort to while deploying TGREP.

Real-time applications such as Voice over IP, audio and video streaming require quality of service (QoS). Such applications are being executed over the public Internet. Since today's Internet largely supports best effort traffic, QoS routing in the best effort environment is required to support real-time applications. Some QoS routing uses source routing and others use flooding of some QoS attributes of the nodes. There were also some variants of the shortest path algorithm reported in the literature. But those algorithms require changes to the packet forwarding engine and logic for loop detection. We believe a better way of implementing QoS routing is to localize the QoS routing changes to the region where QoS has deteriorated and choose loop-free alternate paths. We present such an algorithm based on OSPF (open shortest path first) called the LSR algorithm. In the LSR algorithm, congestion notification is limited to neighbors of the congested node and the neighbors try to use alternate next hops to route packets. Alternate LSR next hop is chosen in such a way that it preserves the next hop property of OSPF routing which enables the LSR algorithm to avoid a loop. We present three such methods to choose an alternate LSR next hop and prove that these methods provide loop-free routing. Our simulation results based on the three methods show that on average the LSR algorithm performs better than the OSPF algorithm in terms of delay and jitter.

We study connection-oriented service in an FDDI-ATM-FDDI heterogeneous network for real-time applications. We design and analyze an algorithm for connection admission control (CAC) for such a network. Upon a request of connection establishment, the CAC determines if the worst case delays of the requesting and existing connections can be satisfied given the available network resources. If so, the CAC allocates appropriate network resources to the requesting connection. The process of allocating resources for homogeneous networks (e.g. FDDI-only or ATM-only) may not be applied directly to a heterogeneous network environment (e.g. FDDI-ATM-FDDI network) because heterogeneity adds more complexity to the process. Hence resource allocation in a heterogeneous network needs more careful analysis than its homogeneous counterpart. We propose a CAC algorithm that will, by proper parameter tuning, allocate sufficient but not excessive network resources to the requesting connection in an FDDI-ATM-FDDI network. We show that the system can achieve satisfactory performance with this CAC algorithm. Our approach is compatible with current network standards and hence can be readily used in practical systems.