International audience

Long-Range (LoRa) is one of the main modulation techniques used for Low Power Wide Area Networks (LPWANs). Recently, it has been extended from the sub-GHz band to the 2.4 GHz Industrial, Scientific, and Medical (ISM) band, enabling the construction of a LPWAN that benefits from global interoperability without duty-cycle limitations. However, the coexistence of LoRa with the other wireless technologies of the 2.4 GHz ISM band is a challenging question. In this article, we make the first performance evaluation of the interference between LoRa and Wireless Fidelity (Wi-Fi) transmissions, by analyzing multiple parameters: LoRa channel occupancy, deployment topology, LoRa physical layer parameters, and the frequency channels used. We also perform simulations to extend our experimental results to other configurations. The performance evaluation is achieved using a generic methodology that can be applied to other wireless technologies. Finally, we provide recommendations for the use and deployment of LoRa that will improve its coexistence with Wi-Fi.

LoRaWAN is a widely used wireless communication standard for the Internet of Things that enables the collection of measurement data from numerous monitoring applications (such as smart metering, pollution, and asset tracking). Most research work has been focused on the performance of LoRaWAN as a function of the uplink traffic, as this is the most common scenario. Nevertheless, downlink traffic is a fundamental building block of the LoRaWAN standard, and a crucial part of applications such as smart healthcare, where reliability is extremely important. In this paper, we study the impact of the downlink traffic and different scheduling policies on the performance of LoRaWAN. We investigate: (i) the impact of an optimal schedule for the downlink traffic, (ii) the choice of the physical layer parameters used to send downlinks to end-devices during the second reception window, and (iii) the choice of the gateway that sends the downlinks. Our results show that even when using an optimal schedule, the presence of downlink traffic reduces the capacity of a LoRaWAN network up to 20%. The most limiting factor is the gateway, due to its duty cycle and half-duplex characteristics.

Long Range (LoRa) technology constitutes one of the major enablers of future Internet-of-Things (IoT) applications, such as monitoring of challenged environments and smart buildings. However, crucial issues in the context of multi-gateway LoRa networks have been overlooked. In particular, most existing methods did not consider the stringent constraint of limited number of demodulators at each gateway. Therefore, we devise a gateway selection method for uplink LoRa transmissions, where this limited availability of demodulators is fully considered. We propose an optimization approach based on the auction mechanism, where each IoT device is pre-assigned to a unique gateway, so as to maximize the total amount of demodulated transmissions without redundancy at the network server. Furthermore, a low complexity method is also designed, where devices are partitioned into groups and where auctions are parallelized. Numerical results show that the proposed methods largely outperform benchmark algorithms in terms of the network utility function and sum-rate, while approaching the upper bound performance. The proposed methods are particularly suited to cope with the inherent dynamics of mobile LoRa IoT networks, as highest gains are attained for demodulation latencies in the order of tens to hundreds of milliseconds.11This work was supported in part by the Grants-in-Aid for Scientific Research (Kakenhi 17K06453 and 20H00592) from the Ministry of Education, Science, Sports, and Culture of Japan and by the NII MoU Grant.

Low-power wide area networks (LPWANs) are being increasingly used in Internet of Things applications, including smart city and environmental monitoring, as they enable communications from low-power end-devices to distant gateways. LoRaWAN is the most common protocol for LPWANs, and is able to automatically tradeoff throughput and reliability thanks to an algorithm called Adaptive DataRate (ADR). However, the LoRaWAN standard imposes mobile nodes to disable the ADR. In this paper, we propose a protocol called MADERE (for Mobile ADR) that attempts to adapt the LoRaWAN parameters for mobile end-devices. We show that MADERE performs well compared to the few existing algorithms from the literature, with limited overhead.

LoRa® and LoRaWAN are one of the most common wireless technologies used today in the Internet of Things. If until now LoRa was strictly used in sub-GHz bands, a new version has been released for the 2.4 GHz band that does not have any duty cycle restrictions and that allows the use of higher datarates, which opens the door to the deployment of new applications (such as asset tracking or indoor localization). In this paper, we present the first extensive evaluation of LoRa 2.4 GHz in a typical indoor environment, both in the presence and lack of human and WiFi activities. We run an exhaustive evaluation of all 128 possible combinations of the different LoRa physical parameters (spreading factor, bandwidth and coding rate) and we show that despite the use of higher frequency, LoRa is capable of maintaining a good connectivity throughout the building, similar to what was observed in sub-GHz bands. Still, some configurations were clearly affected by the daily life activities in the building during the working hours of weekdays.

Long-Range Wide Area Networks (LoRaWAN) enable low-power data collection over long distances, and they are thus widely used for Internet of Things applications, despite their limitations to meet some traffic requirements (e.g., reliability). To achieve the quality of service required by the applications, service differentiation is a promising approach which can be provided by network slicing. In this work, we show that related works are either incompatible with the LoRaWAN specifications, or do not isolate traffic, or assume an a priori known and stable traffic. In this paper, we propose a lightweight approach to achieve slicing in LoRaWAN, compatible with the LoRaWAN specifications. Our approach allows to isolate traffic, to protect confirmed traffic, and to deal with unsolicited traffic.

LoRa is one of the most prominent low power wide area network technologies, and enables to interconnect thousands of devices distributed over areas of several square kilometers. However, the limited number of demodulators present in the hardware of LoRa gateways limits LoRa scalability. In this paper, we argue that scalability can be improved by having gateways collaborate, so that they attempt to demodulate different frames. We propose several algorithms in order to measure the benefits of random-based collaboration and deterministic collaboration. Our simulation results show that random-based protocols improve the baseline performance in most setups, while deterministic protocols improve the network performance when the number of gateways is large, and with many demodulators per gateway.

LoRaWAN is a major player for IoT data collection in large areas. In such networks, the uplink is mainly used, especially for data transmission, whereas the downlink is used for control purposes, including physical layer configuration, over the air activation and acknowledgments. In most studies of the LoRaWAN capacity, the uplink and the downlink are supposed to be orthogonal. This assumption seems acceptable and correct because the LoRa physical layer uses an inverse modulation for the uplink and the downlink. In this work, we use a real testbed composed of software-defined radio (USRPs, GNU Radio) to show that this assumption is wrong: the frame delivery rate decreases by up to 20% when simultaneous transmissions occur between the uplink and the downlink. As far as we know, it is the first time that this result is shown.

Recent monitoring applications extensively rely on low-power wide-area networks, such as those provided by LoRa and LoRaWAN, in order to enable end-devices to communicate over long distances. However, in large-scale deployments, the small throughput of LoRa is further reduced due to collisions. In this paper, we propose a new scheme that aims to decode colliding frames. It relies on frequency detection at each symbol frontier. It reduces detection errors when symbols of similar values are superposed. It is compatible with legacy LoRaWAN communications, and only requires modifications at the gateway. Our simulation results show that our proposition is able to decode more collisions than the other protocols from the literature (about twice more for 16 nodes and SF12), therefore increasing the throughput and thus the scalability of the network.

LoRa (Long Range) is a physical layer designed for low-power wide area networks. It is widely used to provide long range connectivity to Internet of Things devices. In order to improve the limited throughput of LoRa, researchers have proposed several collision resolution algorithms. However, a common software framework to compare these algorithms is lacking. In this paper, we propose an open-source framework using GNU Radio, mainly designed to test and compare collision resolution algorithms, as well as physical layer algorithms. Our framework can help optimizing the parameters of algorithms according to channel conditions such as very low signal to noise ratio for instance. We also discuss technical implementation issues of existing collision resolution algorithms. Finally, we show how our framework can be used for either real experiments on USRPs, or for simulations with a large number of nodes.

LoRa is one of the leading technologies for Low-Power Wide Area Networks and the Internet of Things. Collisions in LoRa might cause retransmissions, which negatively impact the network performance and scalability. Several algorithms have been proposed to decode colliding frames under specific conditions. However, there remain indistinguishable frames due to uncertainties in some or all symbols. In this paper, we propose a general algorithm that significantly improves the recovery capabilities of existing algorithms by leveraging the LoRa coding techniques. Simulation results show that our algorithm can significantly reduce the number of the failed decoding of LoRa frames and improve the performance of the network.

Long Range (LoRa) is a protocol that enables low-power wireless communications over long distances for a wide range of IoT applications. Its main drawback is its limited throughput, which is further reduced by the small number of demodulators in the hardware of the gateway. In this paper, we propose to use each demodulator as efficiently as possible. To do so, we reuse them for short frames during the preamble of long frames. By smartly planning the demodulation of multiple frames, the proposed method enables a recursive reuse of each demodulator. Compared to the benchmark packet arbiter policy, our method is shown to offer throughput and fairness enhancements even with a large number of users, thereby improving the scalability of LoRa systems. Our simulation results show that when the number of nodes is large, 6.5% more frames are decoded and the rate fairness among nodes is improved by 11%.

LoRa is becoming widely used in low-power wide area networks as it enables a communication range of several kilometers with low energy consumption, but with a low bitrate. Collisions in LoRa further reduce the overall performance of the network, and more specifically the throughput. In this paper, we propose a slotted MAC protocol that enables the decoding of colliding LoRa signals. It is based on a new decoding technique at the physical layer that is able to decode the symbols of many frames in collision. Simulation results show that our MAC protocol significantly increases the achievable performance of LoRa networks. For instance, for 25 nodes having a duty-cycle of 10% and with SF7, the throughput with our protocol is 11% larger than the existing protocols.

In LoRa (Long Range), when a collision occurs in the network, each end-device has to retransmit its colliding frame. This reduces the throughput, and increases the energy consumption of the end-devices and the delay of the frames. In this paper, we propose an algorithm to decode colliding synchronized LoRa signals and thus improve the overall performance of the network. Indeed, we use successive transmissions of bitmaps by the end-devices to determine the correct symbols of each colliding frame, instead of retransmitting the whole frames. Simulation results show that our algorithm is able to significantly improve the overall throughput of LoRaWAN, and to decrease the energy consumption and the delay of the transmitters.

The LoRa physical layer is one of the most promising Low Power Wide-Area Network (LPWAN) technologies for future Internet of Things (IoT) applications. It provides a flexible adaptation of coverage and data rate by allocating different Spreading Factors (SFs) and transmit powers to end-devices. We focus on improving throughput fairness while reducing energy consumption. Whereas most existing methods assume perfect SF orthogonality and ignore the harmful effects of inter-SF interferences, we formulate a joint SF and power allocation problem to maximize the minimum uplink throughput of end-devices, subject to co-SF and inter-SF interferences and power constraints. This results into a mixed-integer non-linear optimization, which, for tractability, is split into two sub-problems: firstly, the SF assignment for fixed transmit powers, and secondly, the power allocation given the previously obtained assignment solution. For the first sub-problem, we propose a low-complexity many-to-one matching algorithm between SFs and end-devices. For the second one, given its intractability, we transform it using two types of constraints' approximation: a linearized and a quadratic version. Our performance evaluation demonstrates that the proposed SF allocation and power optimization methods enable to drastically enhance various performance objectives such as throughput, fairness and power consumption, and that they outperform baseline schemes.

Low-Power Wide-Area Network (LPWAN) based on LoRa physical layer is envisioned as one of the most promising technologies to support future Internet of Things (IoT) systems. LoRa provides flexible adaptations of coverage and data rates by allocating different Spreading Factors (SFs) to end-devices. Although most works so far had considered perfect orthogonality among SFs, the harmful effects of inter-SF interferences have been demonstrated recently. Therefore in this work, we consider the problem of SF allocation optimization under co-SF and inter-SF interferences, for uplink transmissions from end-devices to the gateway. To provide fairness, we formulate the problem as maximizing the minimum achievable average rate in LoRa, and propose a SF allocation algorithm based on matching theory. Numerical results show that our proposed algorithm enables to jointly enhance the minimal user rates, network throughput and fairness, compared to baseline SF allocation methods.

The Long Range Wide Area Network (LoRaWAN) standard was mainly developed to meet the requirements of many emerging Internet of Things (IoT) applications as it provides low power and low cost connectivity over long distances. In LoRaWAN, when an end-device transmits an uplink message, gateways that receive this message transmit it to the network server, which, in turn, selects a single gateway to reply to the end-device. LoRaWAN does not specify how to select the gateway. In this paper, we focus on the gateway selection for downlink communications in LoRaWAN in order to improve the throughput of the network. We aim to present and evaluate several algorithms for selecting the best gateway for downlink while increasing LoRaWAN throughput for different types of gateway deployment. We show that the system throughput depends on this deployment and that balancing the number of end-devices per gateway improves the performance compared to choosing the gateway with the highest signal quality.

Long-range low-power wireless communications, such as LoRa, are used in many IoT and environmental monitoring applications. They typically increase the communication range to several kilometers, at the cost of reducing the bitrate to a few bits per seconds. Collisions further reduce the performance of these communications. In this paper, we propose two algorithms to decode colliding signals: one algorithm requires the transmitters to be slightly desynchronized, and the other requires the transmitters to be synchronized. To do so, we use the timing information to match the correct symbols to the correct transmitters. We show that our algorithms are able to significantly improve the overall throughput of LoRa.

Wireless Multimedia Sensor Network (WMSN) can be well suited in precision agriculture especially in rice fields to increase productivity and reduce risks. Real time data are collected from scalar, audio or video sensors to monitor the rice fields and to help farmers in the rice cultivation cycle. In this paper, a new clustering based routing protocol and an integer linear program was proposed, that balances the number of nodes in clusters. It aimed to balance the number of members in each cluster in order to reduce intra-cluster congestion as well as the number of congested cluster-heads and provide efficient multimedia data. Simulation results indicate that our proposed scheme outperforms other protocols proposed in the literature in terms of lifetime, number of members per cluster and the forwarded packets per CH.

In networks that operate during a long time, the routing protocol might have to be changed: this is the case when the network administrator plans a router change. Loop-free transition algorithms are used to ensure that there is no loop during the migration from the initial routing protocol to the final routing protocol. In this paper, we propose a distributed loop-free transition algorithm, called DLF (distributed loop-free heuristic). The algorithm is based on the fact that routing loops resulting from the removal of a node are localized, and can be detected efficiently. We show through simulations that DLF compares well with the existing centralized algorithms, and outperforms the existing distributed algorithm, in terms of migration duration.

Wireless sensor networks have long been characterized by short range communications. Lately, long range wireless sensor networks such as LoRaWAN (long range wide area network) have emerged, increasing the communication range at the cost of reduced bit rate. This paper focuses on the MAC layer of LoRaWAN and studies communications from a gateway to an end-device. We propose a Markov chain model to estimate the delivery delay of an acknowledged communication (with retransmissions) and study the impact of the number of channels, the data-rate and the number of nodes.

In networks that operate during a long time, the routing protocol might have to be changed (in order to apply a routing protocol update, or to take into account a change in the routing metrics). A loop-free transition algorithm has to be used in order to perform the transition to the new routing protocol without generating transient routing loops. In this paper, we propose a loop-free transition algorithm called ACH (avoiding cycles heuristic), which is able to perform the transition in a very small number of steps. Compared to other algorithms of the literature, ACH yields a number of steps which is independent of both the number of nodes and the number of destinations, and thus allows the transition to be performed in a small time. We show through simulations that ACH significantly outperforms other heuristics of the literature, due to its capability to deal with several destinations at once, and due to a priority-based procedure to avoid cycles.

Wireless Sensor Networks (WSNs) are a promising technology to monitor distant or inaccessible areas. As nodes have a limited energy supply, many routing protocols are based on a clustering mechanism: some nodes are elected as cluster heads and have to deal with most of the communication burden of the network, while the other nodes perform only simple operations. In this paper, we propose a new election mechanism with important features: it ensures that all nodes are in range of a cluster head while keeping the number of cluster heads low, it takes into account the residual energy of nodes, and it requires a small communication overhead. We compare the performance of our election mechanism with an optimal election, as well as with the election mechanism of LEACH, which is the main clustering algorithm for WSNs.

Precision agriculture (PA) has been part of one of the most important changes in agriculture where the satellite farming introduced a farming management based on field monitoring. Precision agriculture has been targeted intensive farmlands. Small and medium farms have been forgotten. For this type of farming, it is important to use techniques and organisation that improve cultivation performance at low cost. Personal observation collectors and collective intelligence facilitate effective cultivation monitoring, customising the use of precision agriculture for small and medium farmlands. This paper defines an observation object named warncons (Warning Contents) and proposes an algebraic model to reduce the semantic gap in agriculture early warning management.

In the past few years, energy conservation has been the main focus of researchers working on wireless sensor networks. One of the main technique to save energy is to deactivate periodically the radio module of sensor nodes: nodes alternate periods of activity and periods of inactivity, which is referred to as their duty-cycle. In this paper, we focus on asynchronous duty-cycle mechanisms, as these mechanisms are usually simple, do not require time synchronization and support network changes. We propose an asynchronous MAC protocol based on blind rendez-vous and random wake-up. Our protocol is based on a modification of the IEEE 802.15.4 standard, where activities start at a random time within each activity cycle. Our simulations show that our protocol can achieve a good performance under various scenarios, for small duty cycles (ranging from 0.1% to 5%).

In wireless sensor networks (WSNs) applications such as environmental monitoring, it is essential to design protocols that are energy efficient and that are scalable. The main technique to reduce energy consumption is to use a duty-cycle, where nodes periodically go to sleep. However, the implementation of such a duty-cycle often requires nodes to be synchronized, which is difficult to achieve in a scalable manner. In this paper, we study the delay of blind rendez-vous for unsynchronized nodes, that is the delay required for two unsynchronized nodes to be active at the same time. There are several major advantages of blind rendez-vous: nodes do not need to share a prior knowledge, nodes do not need to be synchronized, and the support for network dynamicity (due to node mobility or failure) is simplified. We show through simulations that the delay to achieve a blind rendez-vous is reasonable in many scenarios.

Wireless sensor networks can accommodate multiple applications by using a multi-stack architecture in order to deliver a large number of QoS. Multi-stack architectures can be optimized by allowing packet exchanges between stacks. However, routing loops may appear because of these exchanges. In this paper, we highlight the problem of routing loops generated when the same packet is routed according to two routing protocols. We define the delayable property of routing protocols by considering that some nodes might hold packets in order to avoid loops in the network. We show that minimizing the number of such nodes is an NP-complete problem. Then, we propose two heuristics to address this issue: a centralized deterministic heuristic requiring a global knowledge of the network, and a distributed stochastic heuristic reducing the number of hops from source to destination. Our two heuristics show important benefits: we reach a gain of up to 67% for the first heuristic and of up to 53% for the second heuristic, in terms of number of hops.

Wireless sensor networks are often deployed for monitoring purposes: when nodes detect the occurrence of a significant event, they transmit an information to a control entity in a multi-hop fashion. When data rate increases, congestion becomes a fundamental issue, especially when an emergency situation generates alarm messages originating from a specific area of the network. Indeed, congestion increases delays and packet losses, and yields to an unfair use of the energy of nodes. In this paper, we propose to improve the ZigBee routing protocol, aiming at reducing traffic congestion. The proposed solution uses intermediate nodes, denoted as pivots, which are selected by the data sources in order to reduce congestion on paths. Simulation results highlight the significant improvement achieved in terms of packet losses and average delays, with respect to the ZigBee routing protocol, while the overhead generated in the network is maintained under control. A mathematical model to derive the average path length and the number of pivots is also provided.

In wireless sensor networks, saving energy is crucial in order to increase the network lifetime. Energy is often saved by synchronizing the nodes activity, and having long periods of inactivity, or by having nodes exchange a global activity schedule. The synchronization and the exchange of a global schedule are two examples where information is boadcast from a specific node to the whole network. In this paper, we focus on the delay required to broadcast information in the whole network using a tree topology. We first show that the diffusion delay can be significantly reduced by utilizing the parallelization of node processing. We provide an algorithm in order to find optimal solutions when transmissions are sequential. Then, we propose a linear algorithm that is able to find good solutions. We compare the exact solution to the heuristic solution on a workstation and conclude that our heuristic is very competitive and can be used to reduce the diffusion delay of a broadcast frame in a tree.

Wireless sensor networks deployed nowadays are traditionally mono-stack: they are operating according to a single combination of one MAC protocol and one network (NWK) protocol. This work proposes a new multi-stack architecture in which several combinations of MAC and NWK protocols are used. This can be achieved by dividing time into time intervals and activating different combinations during each period. Simulations of this approach prove that QoS is mitigated among the combinations. To alleviate the complexity of the time-intervals dimensioning, a formal description of a queue exchange algorithm that allows frames from a time-interval to be sent during another time-interval is proposed. This algorithm improves significantly the global performance of the network.

Several industrial applications require mobile robots to exchange data between them or with a static sink. In this paper, we consider a fleet of mobiles moving in an environment, carrying sensors and actuators. As they move close to a manager, the mobiles exchange data with this static entity. This communication is performed on a meeting channel, different from the channel used for the intra-mobile communications, in order to avoid interferences and to reduce contention. We study how nodes carried by mobiles can associate to the manager network. We show that the association procedure of the IEEE 802.15.4 standard is not adapted to this scenario, because it requires too much time and reduces the time remaining for the communication. Thus, we propose a new association procedure which enables a fast association of nodes. Our results allow us to build an efficient architecture for a wireless personal area network constituted of mobile sensors and static managers.

In wireless sensor networks, energy efficiency is achieved by making nodes sleep. In this paper, we propose SERENA, a node activity scheduling algorithm based on three-hop coloring. We study the performance of SERENA by simulating the conditions of a realistic environment including unidirectional links. The existence of unidirectional links causes color conflicts, which can be solved by exploiting a collaboration with the MAC layer. Such cross layering approach takes into account the real wireless environment and so enables substantial performance improvements (e.g. better reactivity to topology changes). Finally, we evaluate the benefits brought by this solution, namely the increase of network lifetime and spatial reuse.

Slotted CSMA/CA is an algorithm proposed by IEEE 802.15.4 in order to deal with the contention of energy constrained nodes. Slotted CSMA/CA is used during a limited time interval, included in a superframe. Two specific mechanisms defer the transmissions that would occur towards the end of the contention period. Multiple models of the average delay or throughput of slotted CSMA/CA have been proposed recently, but they do not take into account the deference mechanisms. In this paper, we show that the deference mechanisms occurring at the end of the superframe have a significant impact on the performance of slotted CSMA/CA. We prove this by giving the distribution of the delay in a simple scenario, obtained by simulation. Then, we compute the average delay tMACof traversing the MAC layer, as the MAC parameters or the frame size vary. We show that for either reactive WPANs and energy-efficient WPANs, the impact of the deference mechanisms is important.

Wireless sensor networks are increasingly used for remote monitoring, fire detection, emergency response. Such networks are equipped with small devices powered by batteries and designed to be operated for years. They are often based on the ZigBee standard which defines low power and low data rate protocols. As network size and data rates increase, congestion arises as a problem in these networks, especially when an emergency situation generates alarm messages in a specific area in the network. Indeed, congestion occurs as the alarms converge to a specific destination, which results into packet losses and higher delays. In this paper, we propose a solution for congested links, called the PiRAT (Pivot Routing for Alarm Transmission) protocol. It is based on multi-path routing in order to add some diversity in routing the alarms. PiRAT uses intermediate nodes as pivots to reach the destination. Simulation results show that PiRAT has better performance than previous protocols in terms of packet loss, end-to-end delay, congestion and node overload.

Road congestion and traffic-related pollution have a large, negative social and economic impact, and we believe many of these problems can be reduced through investment in monitoring, distribution and processing of traffic information. This paper outlines how our on-going work on the TIME project (transport information monitoring environment) provides a solution, using traffic sensor systems and the design and development of an open and decentralised software framework. We also discuss how we address the privacy and security implications of the increased use of sensors and data processing.

Wireless sensor networks present significant opportunities for fine-grained and continuous monitoring of road traffic, enabling careful city planning, automated road maintenance and accident detection. Users are typically willing to tolerate a small error in car-flow data, in order to reduce the cost of data propagation from the sensor nodes to the gateway nodes, to which users are connected. In this paper, we first examine the relative performance of Fourier- and wavelet-based algorithms for compressing traffic data locally at the sensor nodes. Using real traffic information from the city of Cambridge (UK), we then demonstrate that car-flow data collected across geographically dispersed sensor nodes exhibit strong spatial and temporal correlations. We then combine lossy Fourier-compression with correlation-based compression to achieve further communication savings within a user-specified error threshold. For a tolerated error of 5-15 cars per 5 min, it is shown that exploitation of temporal correlations yields 14-30% savings relative to Fourier compression alone, whilst use of spatial correlations results in 10-35% savings.

In this work, we investigate the nature of spatio-temporal correlations in an urban traffic scenario, and how they can be exploited to reduce the cost of sensor data propagation to the gateway nodes, where users are connected. We conduct experimental analysis of our proposed algorithms using a real dataset of road traffic data generated in the city of Cambridge.

Multicast tree aggregation is an efficient proposition that can solve the multicast forwarding state scalability problem. Existing works on tree aggregation have focused on developing and simulating protocols that build trees dynamically. However, the underlying problem of the impact of the tree construction algorithm on the performance of the protocols remains untouched. In this paper, we propose a study on the number of trees that need to be configured in a domain depending on the tree construction algorithm. We ran extensive simulations on several real domains and with different tree construction algorithms. Our results show that for a given set of multicast groups, even when this set includes all the possible groups, the number of trees that need to be configured is small. This allows a network administrator to configure off-line all these trees in order to maintain a stable set of trees and to have knowledge of the routes used by the multicast packets. Knowing the set of all the possible trees is also useful to determine the best subset to configure and to give an upper bound of the number of different trees.

We present a novel approach to processing continuous aggregate queries in sensor networks, which lifts the assumption of tree-based routing. Given a query workload and a special-purpose gateway node where results are expected, the query optimizer exploits query correlations in order to generate an energy-efficient distributed evaluation plan. The proposed algorithms, named STG and STS, identify common query sub-aggregates, and propose common routing structures to share the sub-aggregates at an early stage. Moreover, they avoid routing sub-aggregates of the same query through long-disjoint paths, thus further reducing the communication cost of result propagation. In this poster, we provide examples to illustrate the functionality and the communication savings of STG and STS compared to the existing tree-based approach.

Recently several multicast mechanisms were proposed that scale better with the number of multicast groups than traditional multicast. These proposals are known as small group multicast (SCM) or explicit multicast (Xcast). Explicit multicast protocols, such as the Xcast protocol, encode the list of group members in the Xcast header of every packet. If the number of members in a group increases, routers may need to fragment an Xcast packet. Fragmented packets may not be identified as Xcast packets by routers. In this paper, we show that the Xcast protocol does not support the IP fragmentation. We show also that avoiding fragmentation limits the group size that can be handled by the Xcast protocol. First, we describe the Xcast protocol, the Xcast+ protocol (which is an extension of Xcast) and we compare these two protocols with traditional multicast protocols. We propose then a generalized version of the Xcast protocol, called GXcast, intended to permit the Xcast packets fragmentation and to support the increasing number of members in a multicast group. The behavior of the GXcast protocol is analyzed according to several criteria. Finally, we present and evaluate with simulations an improvement to GXcast and we conclude that GXcast is a feasible and promising protocol.