This paper presents SFIxM, a flexible evolution of the LoRa modulation scheme designed to facilitate the allocation of transmission parameters across nodes within a gateway’s coverage area. SFIxM increases the number of available quasi-orthogonal channels and, unlike standard LoRa, offers tunable robustness to noise and interference. Because SFIxM introduces a new requirement for time-axis alignment between transmitter and receiver, we present a practical procedure to achieve this synchronisation, along with an analytical estimate of its failure probability, which we find negligible. Based on this foundation, we show that SFIxM provides traffic capacity gains of 40 to 55%, comparable to the gains obtained with receiver diversity, a well-known approach to improving coverage and range, which also turns out to combine well with SFIxM .

Long-Range Frequency Hopping Spread Spectrum (LR-FHSS) is a recent IoT transmission technology that aims at providing wide radio coverage and high data collection capacity, as a single gateway would be able to collect traffic from tens of thousands of devices. LR-FHSS typically targets satellite IoT services or wide-area coverage in the FCC region. To quantify its traffic collection capacity, we propose a detailed analytical model of LR-FHSS and also run extensive simulations. The outcome is that, first, the maximum utilization of the LR-FHSS network is significantly larger than previously reported, even when maxing out the effect of interference between transmissions on neighboring subchannels. Second, the model allows to readily derive the optimal number of LR-FHSS header replicas for obtaining the best goodput; this number is often larger than the limit found in the standard.

\beginabstract In the context of the Factories at Major Accident Risk (FMAR), we consider a scenario consisting of multiple sensor nodes that detect dangerous conditions and raise alarms over a LoRaWAN network. Upon event detection, a large number of nodes try to transmit concurrent alarm messages and the reception of at least one message is sufficient for the central server to react. We propose a scheme in which nodes operate in a slotted-time mode after the detection of the triggering event and they can only transmit one packet in a randomly chosen slot. The choice of the transmission slot follows a specific probability distribution that maximizes the probability of successful reception of at least one packet subject to real-time latency constraints. We provide an optimization framework to find the optimal distribution for choosing a transmission slot. We validate the proposed scheme by simulations and provide numerical results to compare the performance for three different slot choice distributions: i) uniform, ii) the distribution used in the Sift protocol \cite Tay2004, and iii) the proposed optimal distribution. The results show that the proposed optimal distribution leads to much better probability of successful packet delivery than other distributions. \endabstract

This article explores the capacity limits and the tradeoff between the antagonistic means of enabling reliable data delivery in a loaded LoRaWAN cell. In fact, channel attenuation and variability call for robust transmission settings but the associated load increase causes more collisions between frames. In addition to the physical-layer parameters of the LoRa modulation, this article considers the benefits and tuning of interpacket error correction codes (ECCs), which also trade transmission redundancy for reliability. We thus start by proposing a refined packet delivery ratio (PDR) model that improves the ones found in the literature in that, first, it takes into account the dependency between overcoming ambient noise and dominating colliding frames, and second, it considers the sum of interference powers when multiple colliding frames are present, even if the interference preexists. Moreover, the model extends to the case of a gateway with receiver diversity. In a second step, the model allows to set out the level of redundancy at which ECC is the most effective without hindering capacity: a coding rate of one third. When this is fixed, it allows to define the transmission parameters allocation within a cell and thus the size of the cell. We finally develop an ad-hoc Python discrete-event simulator, freely accessible, to complement the model for assessing the effect of inter-SF and near–far interference and to show the benefits of power control in this respect.

This paper considers a LoRaWAN cell in which devices access the channel using the unslotted ALOHA protocol. We propose a model of this access method that combines the effects of collisions with channel fading, by which reception may get buried in noise. Unlike the existing models of LoRaWAN,our model takes into account the frame arrival timing: it distinguishes, on the one hand, the interference created by earlier transmissions with respect to the frame of interest, and on the other hand, the interference by the frames arriving later on. From the results of the model, we draw three observations regarding the improvement of Packet Delivery Ratio (PDR). First, it puts back under the spotlight the often overlooked fact that repeating frames is always beneficial when the desired PDR is above 60%, even though the extra packet transmissions create more collisions. Second, as soon as the node density becomes notable and collisions have a similar impact on losses as attenuation, adding a smaller spreading factor SF6 modulation into the cell list of transmission parameters allows increasing the coverage range. Third, the model shows that cell capacity sometimes grows with the distance to the gateway or with decreased node transmission power, a trend seldom observed in wireless networks.

The Aloha-like access method is a major limiting factor in LoRaWAN networks, with only 18% of channel utilization at best. It also causes high packet losses when node deployment becomes more massive.In this paper, we propose a technique to enhance the reception process of the LoRaWAN gateways, so that channel utilization can reach values up to 35% in a single LoRaWAN cell.We investigate the benefits of concurrent and preemptive reception at the gateways: the capture effect allows to receive a frame even if it collides with a later frame, whereas Message in Message (MIM) reception allows the gateway to drop the current reception and switch to a new more powerful frame. An implementation in the NS-3 simulator allows us to assess the gains of this approach through extensive simulations.

We propose an optimization model for single-cell LoRaWAN planning which computes the limit range of each spreading factor (SF) in order to maximize the minimum packet delivery ratio (PDR) of every node in the network. It allows to balance the opposite effects of attenuation and collision of the transmissions and guarantee fairness among the nodes. We show that our optimization framework improves the worst PDR of the nodes by more than 13 percentage points compared to usual SF boundaries based on SNR threshold. A study of the tradeoff between precision and resolution time of the model shows its effectiveness even with a small number of possible distance limits, and its scalability when the node density increases.

In this paper, we consider the problem of evaluating the capacity of a LoRaWAN cell. Previous analytical studies investigated LoRaWAN performance in terms of the Packet Delivery Ratio (PDR) given a number of devices around a gateway and its range. We improve the model for PDR by taking into consideration that the following two events are dependent: successful capture during a collision and successful frame decoding despite ambient noise. We consider a realistic traffic model in which all devices generate packets with the same inter-transmission times corresponding to the duty cycle limitation at the highest SF, regardless of the distance to the gateway. Based on the developed model, we optimize the Spreading Factor (SF) boundaries to even out PDR throughout the cell. We validate the analytical results with simulations, compare our model with previous work, and experimentally validate the hypothesis of Rayleigh fading for the LoRa channel. The important conclusion from our results is that a LoRa cell can handle a relatively large number of devices. We also show that there is practically no inter-SF interference (cross interference between transmissions with different SFs): interference from higher SFs comes from nodes located farther away, so they face greater attenuation and thus, they do not interfere with lower SF nodes.

Quick and accurate estimation of link quality, and more specifically packet loss probability, is the key element for efficient and effective communications in wireless multi-hop networks. We focus on IEEE 802.15.4 and we posit that losses only occur when noise and interference last long enough and are strong enough relatively to the received signal, to hinder packet reception. So, the key information for any ordered node pair is the signal to noise plus interference ratio distribution, which we obtain by combining the observed noise plus interference power at the receiver with the received signal strength. In this paper, we propose two novel schemes for the estimation of PER: Burst-NISI and Sample-NISI. Burst-NISI is based on high frequency measurements of the power level of ambient noise and interference around a given node. Sample-NISI sporadically samples the power level of ambient noise and interference when the radio operates according to a duty cycle. Using a large scale experimental platform, we show that our packet error rate estimation schemes are accurate for any packet length and diverse experimentation sites with different settings, for which the prediction is within 10 percentage points of the PER value measured a posteriori.

In this paper, we propose to optimize the LoRaWAN®Adaptive Data Rate algorithm in case an inter-packet error correction scheme is available. We adjust its parameters based on analysis of the LoRa channel, supported by real-world traffic traces. The resulting protocol provides high reliability even over low quality channels with comparable Time on Air and similar downlink usage as the LoRaWAN solution. Simulations and emulation fed by real-world channel traces corroborate the analysis.

We propose to optimize the LoRaWAN Adaptive Data Rate algorithm in case an inter-packet error correction scheme is available. We adjust its parameters based on the analysis of the LoRa channel with multiple reception gateways, supported by real-world traffic traces. The resulting protocol provides very high reliability even over low quality channels, with comparable Time on Air and similar downlink usage as the currently deployed mechanism. Simulations corroborate the analysis, both over a synthetic random wireless link and over replayed real-world packet transmission traces.

In this paper, we present the results of extensive experiments on a testbed in the The Things Network (TTN), a public LoRa network. We evaluate the transmission quality of LoRa links by measuring the Packet Reception Rate (PRR) as a function of the payload length. The results show that there is only a slight impact of the payload length on PRR, which means that the bit error rate does not strongly influence the probability of packet reception. Our measurements show that the LoRa channel behaves like a slow fading Rayleigh channel, which translates into probability Ps of being (or not) in a favorable condition for each frame reception: once the frame preamble is received, there is great chance that the whole frame is correctly received. Probability Ps depends on the Spreading Factor and the Signal to Noise Ratio, and often becomes a dominant factor of successful reception depending on the signal strength at a gateway.

We propose a theoretical framework for maximizing the LoRaWAN capacity in terms of the number of end nodes, when they all have the same traffic generation process. The model optimally allocates the spreading factor to the nodes so that attenuation and collisions are optimized. We use an accurate propagation model considering Rayleigh channel, and we take into account physical capture and imperfect SF orthogonality while guaranteeing a given transmission success probability to each served node in the network. Numerical results show the effectiveness of our SF allocation policy. Our framework also quantifies the maximum capacity of single cell networks and the gain induced by multiplying the gateways on the covered area. We finally evaluate the impact of physical capture and imperfect SF orthogonality on the SF allocation and network performances.

Although LoRaWAN offers many advantages, one of its main limitation is the insufficient downlink capacity in large-scale scenarios. This makes reliable communication impractical as, in LoRaWAN, reliability is achieved through the acknowledgement of confirmed messages. The contribution of this paper is to provide a comprehensive description of the problems associated with the presence of downlink traffic and to propose solutions to enable the use of acknowledgments for low to medium downlink load. The proposed mechanisms build on a more adequate and flexible choice of the gateway for each downlink transmission by the network server. Thus, they are entirely compatible with existing LoRaWAN technologies and deployments. Our simulation results show that the proposed solutions enable the successful realization of a higher percentage of LoRaWAN confirmed messages. Thus, the common practice of strongly discouraging the use of confirmed messages could be reconsidered to allow applications that require reliable communication.

All existing models for analyzing the performance of LoRaWAN assume a constant density of nodes within the gateway range. We claim that such a situation is highly unlikely for LoRaWAN cells whose range can attain several kilometers in real-world deployments. We thus propose to analyze the LoRa performance under a more realistic assumption: the density of nodes decreases with the inverse square of the distance to the gateway. We use the LoRaWAN capacity model by Georgiou and Raza to find the Packet Delivery Ratio (PDR) for an inhomogeneous spatial distribution of devices around a gateway and obtain the number of devices that benefit from a given level of PDR. We analyze the LoRaWAN capacity in terms of PDR for various spatial configurations and Spreading Factor allocations.

In the DNS resolution process, packet losses and ensuing retransmission timeouts induce marked latencies: the current UDP-based resolution process takes up to 5 seconds to detect a loss event. We find that persistent DNS connections based on TCP or TLS can provide an elegant solution to this problem. With controlled experiments on a testbed, we show that persistent DNS connections significantly reduces worst-case latency. We then leverage a large-scale platform to study the performance impact of TCP/TLS on recursive resolvers. We find that off-the-shelf software and reasonably powerful hardware can effectively provide recursive DNS service over TCP and TLS, with a manageable performance hit compared to UDP.

Data Pendulum [1] designates the packet backlog switch from one side of a bottleneck link to the other in presence of two antiparallel TCP connections. This paper completes the explanation of the Data Pendulum dynamics by elucidating on which side the backlog grows, while identifying the relevant parameters in the process. We also explain why and when the data backlog may switch sides without any packet loss. We propose a model for the dynamics of antiparallel TCP connections that effectively explains the throughput observed in real conditions. Moreover, the model is applicable to a more realistic case in which buffers are dimensioned in packets rather than bytes, the case that previous models could not capture.

Reliable wireless communication even in adverse conditions is the key for building the energy efficient and dependable Internet of Things. In this paper, we explore the benefits of channel diversity for enabling efficient wireless communication: we propose MRR (Multi-channel Round-Robin), a backward-compatible evolution of beacon-enabled IEEE 802.15.4 in which energy constrained nodes take advantage of additional active periods operating on different channels in a round-robin way. Each active period starts with a beacon sent on a cyclically changing channel, which then allows an associated device to transmit data on the channel used for the beacon. MRR schedules the additional active periods at carefully selected instants to avoid direct beacon collisions. To motivate our work, we first experimentally corroborate previous findings that channel diversity is an effective way of mitigating variable or poor transmission conditions. Then, we observe that channel diversity improves the quality of transmission even better than expected - it appears that wireless sensor nodes have a radiation pattern that changes significantly from one frequency channel to another, which often results in a considerably improved gain when using the right communication channel. The evaluation of the MRR scheme through measurements on a real indoor multihop testbed shows that the proposed scheme results in significantly improved Packet Reception Ratio even without resorting to e.g. channel blacklisting. These results confirm the benefits of multichannel operation and exhibit a fully functional solution that does not add a large overhead compared to using a single channel.

We consider the crucial problem of routing in wireless sensor networks. Routing protocols need to deal with topology changes while keeping the routing overhead low, especially the number of broadcasts, to save energy. In this paper, we consider the problem of building and adapting default routes for convergecast and host routes for downward traffic. We first propose a local repair scheme that allows arbitrarily long periods for rebuilding the tree/DODAG of default routes. Second, we design a scheme for preventing packet forwarding along routing loops, which may occur in the combination of broken host routes and default routes. We achieve this goal without adding an additional header to packets. The two schemes along with on-demand host route construction provides a complete routing solution that addresses many of the shortcomings of RPL. We validate the proposed schemes with Cooja emulations and an experimental evaluation on a real-world sensor network testbed.

This paper presents GreenNet, an energy efficient and fully operational protocol stack for IP-enabled wireless sensor networks based on the IEEE 802.15.4 beacon-enabled mode. The stack runs on a hardware platform with photovoltaic cell energy harvesting developed by STMicroelectronics (STM) that can operate autonomously for long periods of time. GreenNet integrates several standard mechanisms and enhances existing protocols, which results in an operational platform with the performance beyond the current state of the art. In particular, it includes the IEEE 802.15.4 beacon-enabled medium access control (MAC) integrated with lightweight IP routing for achieving very low duty cycles. It offers an advanced discovery scheme that accelerates the process of joining the network and proposes an adaptation scheme for adjusting the duty cycle of harvested nodes to the available energy for increased performance. Finally, it supports security at two levels: a basic standard secure operation at the link layer and advanced scalable data payload security. This paper describes all techniques and mechanisms for saving energy and operating at very low duty cycles. It also provides an evaluation of the performance and energy consumption of GreenNet.

In this paper, we investigate schemes for energy-efficient multi-hop broadcasting in large-scale dense Wireless Sensor Networks. We begin with an initial simplified study of the schemes for relay selection. Our first finding is that MPR-based (Multipoint Relay) mechanisms work poorly in a dense network while the recently proposed Multicast Protocol for Low power and Lossy Networks (MPL) protocol based on Trickle performs better. However, Trickle requires to overhear packet retransmissions in the vicinity, while sensor nodes try to avoid overhearing by periodically waking up and going to sleep to save energy.

IEEE 802.15.4 defines a popular MAC standard for wireless sensor and actuator networks. With the default parameters, under medium to high load, 802.15.4 generates excessive collisions and packet losses. Low duty cycles even exacerbate the problem, because more nodes become active after long periods of sleep and contend for channel access. In this paper, we have applied the models that led to the Idle Sense access method for 802.11 to the 802.15.4 slotted CSMA/CA, taking into account the central role of the coordinator and also the bursty nature of the traffic. Surprisingly, the approach perfectly applies to 802.15.4 even if the principles of the two access methods fundamentally differ. Based on the model, we propose ABE, an adaptation method that adjusts the contention window to optimal values so that the network obtains high throughput along with low duty cycles leading to low energy consumption. The method converges to near-optimal back off values even under bursty traffic and for any number of contending nodes.

In wireless sensor networks, the PHY layers use robust modulations and spread spectrum techniques as in IEEE 802.15.4, which often allows to capture a frame successfully even if it collides with another transmission. In this work, we propose eMiM (Enhanced MiM), a new capture scheme that takes into account both SINR and the collision timing in the switching decision. We have carried out extensive simulations to evaluate the performance gains compared to the MiM scheme. The simulation results show that the proposed capture scheme results in a substantial improvement of the frame reception ratio, as well as fairness.

In IEEE 802.15.4 networks, nodes access the medium using a CSMA/CA method to avoid collisions. As the networks use robust physical layers, they benefit from the capture effect: a receiver can correctly receive a frame even if it collides with another transmission. The capture effect and the access scheme result in significant unfairness between nodes depending on their positions. In this paper, we propose new adaptive medium access control algorithms to improve fairness and performance. The algorithms adapt the contention parameters and transmission power depending on the observed busy channel and transmission failure rates to achieve a target frame delivery rate. We have carried out extensive simulations to evaluate the performance gains of the algorithms compared to legacy IEEE 802.15.4 with static transmission power and default contention parameters. Simulations show important improvement in terms of the frame delivery rate, energy consumption, and fairness.

In this paper, we study current routing issues in Low Power and Lossy Networks (LLNs) and propose two enhancements to the RPL routing protocol. First, we focus on the establishment and maintenance of the Destination Oriented Directed Acyclic Graph (DODAG) for upward traffic and propose a low overhead mechanism that guarantees the absence of loops and quickly reacts to topology changes using a link reversal algorithm. Second, we propose either proactive or on demand establishment of downward routes from the DODAG root to leaf nodes. Simulation results show that our enhancements reduce the convergence time, generate less routing messages, and improve the performance in a network subject to link failures.

We consider extended LANs composed of two different parts: a static one with hosts and routers, and a dynamic one with mobile hosts. We propose a scheme for interconnecting the two parts that builds on OLSR to establish compact routing tables, by aggregating addresses into prefixes whenever possible. Prefix advertisements imply possible conflicts of prefixes covering the same address ranges, thus we propose an extension to OLSR for solving this issue. We report on performance improvement measured on an experimental testbed.

Size-based scheduling improves data transfer response times by favoring flows at an early stage. Although appealing, these techniques raise concerns as they require to keep track of the volume of data sent by each and every ongoing connections and they may starve long-lived flows even if they use up limited bandwidth. Early Flow Discard (EFD) scheduling addresses these issues and we present its adaptation to infrastructure 802.11 networks where the access point downlink queue naturally builds up. To deal with this problem, EFD needs to take into account bi-directional traffic, so that it effectively controls uploads and downloads even though EFD applies to the downlink buffer only. It appears that even with limited buffers, which translates into limited memory of flows for EFD, the most simple flavor of bidirectional EFD -a simple pair of FIFO queues and tracking flow transferred volumes with a packet granularity-enables to rip the full benefit of size-based scheduling, without any of the aforementioned drawbacks.

Wireless LANs suffer from performance problems caused by insufficient medium access opportunity given to the access point. Consequently, the downlink buffer fills up, which often leads to packet losses. We propose to address this problem by using a size-based scheduling approach, which is known to favor short flows and the start up of new ones-a very appealing property from the user's perspective as interactive applications and new flows are serviced quickly. Still, size-based scheduling policies have a well-known Achilles heel: large flows can block each other for long periods of time and low rate multimedia transfers may end up with a low priority when their accumulated transferred volume becomes large. To solve the above deficiencies, we propose a new packet scheduling scheme called Least Attained Recent Service (LARS) that applies a temporal decay to the volume of data associated with each flow. In this way, its priority depends more on what has happened recently. With this strategy, LARS can bound the impact of a new arriving flow on ongoing flows, thus limiting lock out durations. It can also efficiently protect low rate multimedia transfers irrespectively of the load conditions.

We propose a novel view on efficient packet forwarding in wireless mesh networks based on a molecular analogy in which mesh routers are either nuclei or electrons in an atom that corresponds to a 802.11 cell. In this view, a mesh network appears as a collection of spatially distributed 802.11 cells operating on different channels. We define Molecular MAC that uses dynamic channel switching at neighbor mesh routers to efficiently forward packets over multiple hops. To avoid deafness, nuclei notify electrons about pending packets before electrons pull them from nuclei for reception or further forwarding. We evaluate the proposed scheme through simulation and compare with other proposals. Our results show that Molecular MAC obtains much better performance in terms of throughput, packet delivery rate, end-to-end delay, and fairness.

We provide an analytical framework for preamble sampling techniques for MAC protocols in wireless sensor networks, from which we derive closed-form formulas for lifetime and reliability calculations. In addition to take into account transmitter behavior that controls the form and the content of the transmitted preamble, our model also considers receiver behavior that controls the duration of preamble reception in case of successful and failed reception. Along with both transmitter and receiver behavior, our model considers a non-perfect channel and thus takes into account the impacts of transmission errors and retransmissions on lifetime and reliability of preamble sampling protocols. Numerical results show that no protocol is universally optimal; that is, each protocol has its own optimal operation point that depends on the given channel and load conditions.

Many researchers have addressed the problem of QoS differentiation in 802.11 wireless networks, however no method proposed so far benefits from all desirable properties: high aggregate throughput even for a large number of contending stations, fair allocation to all stations in the same class, fast adaptation to changing conditions, and support for absolute priorities. If we consider the IEEE 802.11e standard, its EDCA (enhanced distributed channel access) access method suffers from an increased collision rate when the number of stations increases. In this paper, we propose a novel access method that supports both relative proportional throughput allocation and absolute priorities in 802.11 wireless networks. The method is efficient, scalable, and fair. It builds on the idea of the Idle Sense method that provides the optimal throughput and fairness for 802.11 WLANs [1]: each station adjusts its contention window based on the observed average number of idle slots. We achieve absolute priority differentiation by setting the target value for the number of idle slots to a small value, so that the absolute priority class gains all the available throughput. The method also supports relative proportional throughput allocation in which several classes share the available throughput according to desired ratios. Our simulations show that the proposed method achieves its objectives of relative and absolute differentiation both with respect to the aggregated throughput and the speed of convergence. Unlike 802.11e EDCA, it presents very good scalability - the throughput remains almost constant in function of the number of contending stations.

We propose a packet scheduling discipline called virtual flow queueing (VFQ) to improve the end-to-end performance of TCP connections over IEEE 802.11 WLANs. VFQ is based on an inter-layer approach: the IP layer schedules packets over the wireless link according to the information coming from both the transport and MAC layers. The scheduler computes virtual transmission times that take into account the channel time at the MAC layer needed to transfer both TCP and ACK segments. The scheduling discipline achieves fair allocation of wireless channel capacity to TCP connections in both directions (upload and download). VFQ also supports different weights like in the classical WFQ scheduling so that the access point can for instance give preference to download connections. In this paper, we describe the VFQ implementation in the Linux kernel and present the results of measurements that show very good performance behavior of VFQ compared to the FIFO scheduling discipline currently deployed in 802.11 access points.

In atypical deployment of IEEE 802.11 wireless LANs in the infrastructure mode, an access point acts as abridge between the wireless and the wired parts of the network. Under the current IEEE 802.11 Distributed Coordination Function (DCF) access method, which provides equal channel access probability to all devices in a cell, the access point cannot relay all the frames that it receives on the downlink. This causes significant unfairness between upload and download connections, long delays, and frame losses. This unfairness problem comes from the not-so-complex interaction of transport-layer protocols with the MAC-layer access method. The main problem is that the access point requires more transmission attempt probability than wireless stations for correct operation at the transport layer. In this paper, we propose to solve the unfairness problem in a simple elegant way at the MAC layer. We define the operation of an Asymmetric Access Point that benefits from a sufficient transmission capacity with respect to wireless stations so that the overall performance improves. The proposed method of operation is intrinsically adaptive so that when the access point does not need the increased capacity, it is used by wireless stations. We validate the proposed access method by simulation to compare it with other solutions based on IEEE 802.11e. Unlike many papers in this domain, which only validate MAC-layer modifications through simulation or analytical modeling, we provide measurement data gathered on an experimental prototype that uses wireless cards implementing the proposed method.

very selective A+

International audience

In this paper, we present an evaluation of chosen wireless LAN access methods involving stations with different bit error rates: n-1 stations in ideal transmission conditions (BER = 0) and 1 station with a given bit error rate (BER ne 0). The simulation results show that the IEEE 802.11 DCF and its modifications (Slow Decrease, AOB) are very sensitive to transmission errors, whereas Idle Sense provides good channel access fairness: the value of the contention window is almost the same regardless of transmission errors, so that the throughput difference between stations subject to different bit error rates corresponds only to the proportion of lost frames.

In this paper, we consider the problem of blocked stations that appears in some spatial configurations of multi-hop wireless networks based on the 802.11 DCF (distributed coordination function). The problem leads to starvation of at least one station caused by the presence of neighbor stations within its carrier sensing range that do not hear each other. We propose Forced Transmissions, a simple and efficient solution to this problem. It consists of detecting that a station is blocked by others stations and forcing a transmission. This results in a collision that increases the contention windows of blocking stations and leaves some channel time to the blocked station for transmitting. The blocked station forces transmission only with some probability adjusted in function of the time spent waiting for the channel. Our simulations show that the proposed method increases the minimal throughput of the blocked stations at the expense of a slight degradation of the total throughput, which is unavoidable in any DCF based solution.

In this paper, we address the issue of evaluating performance of wireless LANs in multicell scenarios. We try to understand the complex behavior of the DCF (distributed coordination function) access method defined in the IEEE 802.11 standard [1] and its modifications proposed for improving performance: slow decrease [2], asymptotically optimal backoff [3], and idle sense [4]. We analyze the influence of overlapping cells and large multicell environments on their performance. Our results show that the IEEE 802.11 DCF and its two modifications (slow decrease and AOB) exhibit important unfairness between stations close to the access point and those near the border of a neighbor cell. Idle Sense performs much better: it provides much better fairness than the IEEE 802.11 DCF and its modifications. It also obtains the highest throughput when stations adapt their bit rate to channel conditions.

MAC protocols based on preamble sampling techniques like WiseMAC and BMAC offer significant energy savings for multihop wireless sensor networks. Although preamble sampling based techniques efficiently reduce idle listening in low-rate data networks, they introduce extra overhead due to possible reception of some irrelevant frames. In this paper, we identify the problem of receiving irrelevant frames and propose to replace the continuous preamble by a series of small frames that we call micro-frames. The resulting Micro-Frame Preamble (MFP) scheme makes it possible for a node to switch its radio off to avoid receiving irrelevant frames and improve energy savings. In the paper, we evaluate the theoretical energy savings of MFP compared with traditional preamble sampling techniques. We also discuss some experiences with our implementation of MFP on the Freescale MC13192 SARD.

Previous studies have shown that preamble MAC protocols have higher energy efficiency compared to traditional low power MAC protocols based on common sleep/wakeup schedules. One efficient implementation of preamble MAC protocols is frame preamble MAC in which the MAC preamble is replaced by a series of frames. In this paper, we consider four frame preamble MAC protocols, namely: persistent MFP (micro frame preamble), non persistent MFP, persistent DFP (data frame preamble), and non persistent DFP. We study their energy cost and communication reliability by assuming a simple binary symmetric channel error model

We consider wireless LANs such as IEEE 802.11 operating in the unlicensed radio spectrum. While their nominal bit rates have increased considerably, the MAC layer remains practically unchanged despite much research effort spent on improving its performance. We observe that most proposals for tuning the access method focus on a single aspect and disregard others. Our objective is to define an access method optimized for throughput and fairness, able to dynamically adapt to physical channel conditions, to operate near optimum for a wide range of error rates, and to provide equal time shares when hosts use different bit rates.We propose a novel access method derived from 802.11 DCF [2] (Distributed Coordination Function) in which all hosts use similar values of the contention window CW to benefit from good short-term access fairness. We call our method Idle Sense, because each host observes the mean number of idle slots between transmission attempts to dynamically control its contention window. Unlike other proposals, Idle Sense enables each host to estimate its frame error rate, which can be used for switching to the right bit rate. We present simulations showing how the method leads to high throughput, low collision overhead, and low delay. The method also features fast reactivity and time-fair channel allocation.

We analyze in this paper the fairness of the 802.11 DCF (distributed coordination function) access method. Short-term fairness is an important property of a MAC layer for obtaining short delays. We show that contrary to the common wisdom, a 802.11 cell with two hosts does not exhibit short-term unfairness. We propose a new fairness index: the number of inter-transmissions that other hosts may perform between two transmissions of a given host. By means of this index, we analyze the fairness of 802.11 for the case of two hosts and derive the probability distribution of the number of inter-transmissions. Our results show that even on the short term time scale the 802.11 DCF access method provides fairness to competing hosts. When compared with slotted ALOHA, a multiple access randomized protocol with good fairness properties, 802.11 presents even better fairness. To validate the model, we compare the analytical results with experimental histograms obtained via simulations and measurements.

We consider the problem of interconnecting hosts in spontaneous edge networks composed of various types of wired or wireless physical and link layer technologies. We argue that this kind of networks requires a more sophisticated approach than standard IP forwarding: communication paths should be managed on a per flow basis, multiple paths need to be maintained to cope with link failures or changing topologies, and the interconnection architecture should provide a means for acquiring the information on destination reachability. To experiment with our approach, we have designed and implemented Lilith, a prototype of an interconnection node for spontaneous edge networks. We handle network dynamics by establishing MPLS (multi protocol label switching) label switched paths (LSP) on demand with a reactive ad hoc routing protocol. We present some measurements that show good performance with respect to the standard IP forwarding and important performance gains when multiple paths are used.

We propose to provide QoS differentiation at the IP level based on the DiffServ model over the standard distributed coordination function (DCF) method of 802.11. By constraining the output rate of each host in a cell via DiffServ scheduling and traffic shaping mechanisms, we can keep the 802.11 network in the state of nonsaturation so that the time-sensitive expedited forwarding (EF) class benefits from stable short delays. We analyze the performance of 802.11b to find how we should constrain the rate of DiffServ sources and present extensive measurements that show how the QoS differentiation is achieved.

The performance of the IEEE 802.11b wireless local area networks is analyzed. We have observed that when some mobile hosts use a lower bit rate than the others, the performance of all hosts is considerably degraded. Such a situation is a common case in wireless local area networks in which a host far away from an access point is subject to important signal fading and interference. To cope with this problem, the host changes its modulation type, which degrades its bit rate to some lower value. Typically, 802.11b products degrade the bit rate from 11 Mb/s to 5.5, 2, or 1 Mb/s when repeated unsuccessful frame transmissions are detected. In such a case, a host transmitting for example at 1 Mb/s reduces the throughput of all other hosts transmitting at 11 Mb/s to a low value below 1 Mb/s. The basic CSMA/CA channel access method is at the root of this anomaly: it guarantees an equal long term channel access probability to all hosts. When one host captures the channel for a long time because its bit rate is low, it penalizes other hosts that use the higher rate. We analyze the anomaly theoretically by deriving simple expressions for the useful throughput, validate them by means of simulation, and compare with several performance measurements.

We present a distributed path selection procedure which has the objective of balancing the load in the network. It is used here in a connection oriented framework and is based on selecting, at each hop along a partially computed path, any one of the available routes to the destination. As far as we know, our approach is the first distributed multipath routing algorithm guaranteed to be loop-free after convergence, while at the same time basing its route computation on the actual load on the links of the network, as opposed to using some administrative, static, metrics.