Number Of Retransmissions Research Articles

Markov‐based analysis for cooperative HARQ‐aided NOMA transmission scheme in 5G and beyond

AbstractIn the fifth generation (5G) wireless communication system, the non‐orthogonal multiple access (NOMA) scheme has recently been proposed as a possible candidate for improving spectral efficiency. We look at power‐domain NOMA in this study, in which, multiple simultaneous transmissions are allowed at the same channel with different transmission powers. To achieve reliable transmissions in NOMA, the hybrid automatic repeat request (HARQ) scheme can be integrated into NOMA. The performance of a downlink NOMA system using the Type I HARQ method is investigated in this research. This type of HARQ does not entail expensive device which is considered one of the requirements for a massive 5G‐based Internet of Things network. Unfortunately, in the literature, it has been shown that deriving closed‐form expressions for the outage probabilities of HARQ‐aided NOMA systems is intractable due to the incorporation of multiple fractional random variables. Consequently, in this article, we follow a simple yet efficient methodology to study our considered HARQ‐aided NOMA system. The HARQ‐assisted NOMA system on the downlink is modeled as an absorbing discrete‐time Markov chain (DTMC). Thereafter, under perfect and imperfect channel estimation, we use this model to examine system performance in terms of outage probability and estimated number of retransmissions. Interestingly, we demonstrate that our DTMC‐based analysis is straightforward, repeatable, and accurate. Moreover, in this article, we propose a cooperative HARQ (CHARQ) scheme for the downlink NOMA system. According to CHARQ, the users cooperate with the base station in the retransmission process that will result in improving the performance of NOMA. Monte Carlo simulations have been conducted to not only verify the accuracy of our DTMC‐based analysis, but also manifest the superiority of CHARQ over HARQ.

Blockchain Based Secure Routing and Trust Management in Wireless Sensor Networks.

In this paper, an encryption and trust evaluation model is proposed on the basis of a blockchain in which the identities of the Aggregator Nodes (ANs) and Sensor Nodes (SNs) are stored. The authentication of ANs and SNs is performed in public and private blockchains, respectively. However, inauthentic nodes utilize the network’s resources and perform malicious activities. Moreover, the SNs have limited energy, transmission range and computational capabilities, and are attacked by malicious nodes. Afterwards, the malicious nodes transmit wrong information of the route and increase the number of retransmissions due to which the SNs’ energy is rapidly consumed. The lifespan of the wireless sensor network is reduced due to the rapid energy dissipation of the SNs. Furthermore, the throughput increases and packet loss increase with the presence of malicious nodes in the network. The trust values of SNs are computed to eradicate the malicious nodes from the network. Secure routing in the network is performed considering residual energy and trust values of the SNs. Moreover, the Rivest–Shamir–Adleman (RSA), a cryptosystem that provides asymmetric keys, is used for securing data transmission. The simulation results show the effectiveness of the proposed model in terms of high packet delivery ratio.

ETX-Aware Energy-Efficient Algorithm to Reduce Packet Retransmissions in the Internet of Things

Objectives: An energy-efficient optimum path selection to reduce the number of packet retransmissions in a path is proposed in this paper using a new metric New-Expected Transmission Count (N-ETX), Residual Energy (RE), and Path Objective Function (POF). Methods: Energy efficiency, packet retransmission, and RE is considered while making the best choice of the optimum path. The refined N-ETX metric is employed to measure the Packet Delivery Ratio (PDR) of a node. If the PDR ratio is higher, this node is included in the optimum path set. The proposed algorithm avoids the inclusion of critical nodes such as dead nodes if the RE of the nodes is lower than the predefined Energy Threshold (ETHR). The Unicast message reply to strategy significantly decreases the quantity of control message overhead. The T-test’s Degree of Variance and Degree of Independence is used to discover the energy-efficient optimum path. It is demonstrated through performance outcomes that the obtained Quality of Service (QoS) parameters for the proposed algorithm stands superior to the state-of-art protocols. Findings: An algorithm established on N-ETX and RE metrics is proposed, and the discussions have shown that this approach is energy efficient. This metric decreases the quantity of packet retransmission effectively for a path. As an outcome, the sensor’s higher energy consumption is reduced. The proposed algorithm outperforms the current algorithm by decreasing energy consumption, reducing the quantity of packet retransmission, improving the PDR ratio, and enhancing the network life expectancy. The simulation outcome established that energy consumed is decreased by 19.07%, RE of the nodes is increased by 10.51%, PDR ratio is increased to 98%, and network lifespan is increased by 30.54%. Novelty: Internet of Things (IoT) gadgets are bound by hardware limits regarding computation, memory, and energy proficiency. To effectively transmit data packets, every IoT device must have a routing communications protocol that is easy to create. IoT devices necessitate the demand for self-adaptive routing algorithms. During network operation, packet energy dissipation during the broadcast and reception process is significantly higher when assessed to otherenergy-consuming processes like sensing, data processing, etc. Keywords: LOADng; Energy Efficiency; ETX; Packet Retransmission; Residual Energy

Increasing Reliability on UAV Fading Scenarios

Unmanned aerial vehicles (UAVs) are the next technology to be incorporated into a telecommunications network to improve command and control on a large scale in both line-of-sight (LOS) and non-line-of-sight (NLOS) conditions. However, there is still room for improvement in terms of reliability. This paper investigates Constant Packet Combining (CPC) and Adaptive Packet Combining (APC) techniques applied to Unmanned Aerial Vehicle (UAV) communication in the presence of large-scale fading, where the channels are subject to sudden degradation for long periods due to obstructions. We use Single Carrier (SC) Frequency Domain Equalization (FDE) combined with the Iterative Block Decision-Feedback Equalizer (IB-DFE) to handle command and control messages mapped for UAV use cases. We present closed-form equations for the equalization design as well as the performance parameters such as Bit Error Rate (BER), the Packet Error Rate (PER), the throughput, the retransmissions amount, the goodput (the transmission rate without the retransmissions quantity), and the outage probability. Then, we analyze the system performance using correlated, independent, and equal channels. There is a trade-off between the overall available power, throughput, and reliability. For instance, more retransmissions result in higher reliability, power consumption and lower goodputs (effective data rates). CPC validates the transmission system and confirms the improvement of BER and PER parameters without energy efficiency optimization. APC is appealing because it can reduce the number of retransmissions for all channels used with the advantage of meeting energy efficiency requirements by adapting the overall power to the scenario experienced by the UAV.

Signal coverage analysis with link adaptation in Narrowband-IoT

The aim of the Narrowband-IoT (NB-IoT) technology is deployed to provide massive connections in IoT cellular technologies and this technology was released by the Third Generation Partnership Project (3GPP) in Release-13(R-13). This is one of the secured spectrum Low Power Wide Area Networks (LPWANs) because it provides low energy and conveys the information over a wide area with the authentication process and also, NB-IoT utilized for low complexity IoT devices. The main intention of the NB-IoT is used to extend coverage compared to legacy wireless communication technologies such as WiFi, Bluetooth, Zigbee, etc. Hence, to achieve this extended coverage parameter, alter retransmission numbers, Bandwidth, and Resource Units (RUs) using Signal to Noise ratio, Bandwidth Efficiency, and Energy per Transmitted bits in uplink transmission and also, analyze the impact of transmission time, bandwidth and retransmission numbers for Maximum Coupling Loss (MCL). Moreover, we intend to introduce a link adaptation algorithm for increasing coverage in uplink transmission.

MAC chastised Dynamism Efficient in Wireless Device Lattice Spending Mistralapproach

In Wireless sensor Network, several researchers have provided different routing protocol for sensor networks, particularly routing protocols depending on clusters protocols. Reliability of nodes is necessary parameter in effective sensor networks. We use MAC protocol for controlling the network packets. This is because the usage of cluster based routing has several merits like minimized control messages, re-usability of bandwidth and enhanced power control. Different cluster based routing protocol is proposed by many researchers for the purpose of reducing the consumption energy in wireless sensor networks. Those techniques reduces the energy consumption but with several disadvantages like lack of QoS, inefficient transmission, etc., To overcome those problems, modified QoS enhanced base station controlled in Mistrial Approach (flooding Technique) for wireless sensor networks is proposed in this work. Here we reduce the number of retransmission and detect the overlay packets in networks using proposed approach. Simulation results show the better energy consumption, Maximum Life time & Efficient Bandwidth is achieved by flooding management when compared to the conventional techniques

Energy Adaptive and Max-Min based BFS Model for Route Optimization in Sensor Network

Background: The restricted energy and network life is a critical issue in the real-time sensor network. The occurrence of low-energy and faulty intermediate nodes can increase communication failure. The number of intermediate nodes affects the number of re-transmission, communication-failures, and increases the energy consumption on the routing path. The existing protocols take the greedy decision on all possible intermediate nodes collectively by considering one or more parameters. Objective: This work divides the distance between the source and destination into coverage-specific zones for restricting the hop-count. Now each zone is processed individually and collectively for generating the energy effective and failure preventive route. Methods: In this paper, the energy and coverage weighted BFS (Best First Model) algorithm is presented for route optimization in the sensor network. The max-min BFS is implied on sensor nodes of each zone and identified as the most reliable and effective intermediate node. The individual and composite weighted rules are applied to energy and distance parameters. This new routing protocol discovered the energy-adaptive route. Results: The proposed model is simulated on a randomly distributed network, and the analysis is done in terms of network life, energy consumption, hop count, and the number of route switching parameters. The comparative analysis is done against the MCP, MT-MR, Greedy, and other state-of-art routing protocols. Conclusion: The comparative results validate the significance of the proposed routing protocol in terms of energy effectiveness, lesser route switching, and improved the network life.

Opportunities of hybrid random access protocols for M2M communications in LTE/LTE-A networks

Machine-to-machine (M2M) communications on Long-term evolution (LTE) networks form a substantial part for the Internet-of-things (IoT). The random access procedure is the first step for M2M devices to access network resources. Many researchers have attempted to improve the efficiency of the random access procedure. This work revisits the performance of the hybrid random access protocols which combine congestion control techniques with collision resolution techniques. In particular, we investigate two hybrid protocols. The first one combines the pre-backoff (PBO) with tree random access (TRA), and the second one combines dynamic access barring (DAB) with TRA. The probability analysis is presented for both protocols. The performance is evaluated based on the access success rate, the mean throughput, the mean delay, the collision rate and the mean retransmissions. The simulation results show that the hybrid protocols achieve the highest success rate and throughput with moderate delay and low collision rates with a lower mean number of retransmissions compared to three benchmarks that apply either a congestion control or a collision resolution. The opportunities of future developments of hybrid protocols are listed at the end of this paper to highlight the issues that could be investigated to improve the performance of hybrid random access protocols.

Efficient configuration for multi-objective QoS optimization in wireless sensor network

Wireless sensor network (WSN) is a self-organizing network consisting of many sensor nodes situated in the monitoring area. One of the most important challenges of WSNs is to configure the network in a way to improve the quality-of-service parameters, and thus enhance the network performance and reduce energy consumption. Energy consumption increases due to several reasons such as unsuccessful delivery of packets to the receiver, re-transmission of packets, delay in packet delivery to the sink, low received signal strength, inadequate link quality, noise level, etc. Each sensor network consists of different configuration parameters including the distance between receiver and transmitter, packet arrival time, packet size, maximum queue size, maximum number of transmissions, transmission power level, etc. All the mentioned parameters will affect several indicators such as latency, received signal strength, link quality, successful/fail packet delivery, noise level, and the number of re-transmissions. Finding the relationship between these two sets of parameters and optimizing them reduce energy consumption in the network and increase network lifetime. To achieve this, an efficient and new method based on support vector regression and genetic algorithm called MSOG is presented in this paper to model and optimize wireless sensor network indicators (such as delay, received signal strength, packet arrival time, successful packet delivery). The simulation results show that MSOG has better outcomes in optimizing network parameters and indicators in comparison with a similar algorithm.

Enhanced packet reordering procedure to improve TCP communication

PurposeThe research work is attained by Spurious Transmission–based Enhanced Packet Reordering Method (ST-EPRM). The packet reordering necessity is evaded by presenting random linear network coding process on wireless network physical layer which function on basis of sequence numbers. The spurious retransmission happening over wireless network is obtained by presenting monitoring concept for reducing number of spurious retransmissions because it might need more than three DUPACKs for triggering fast retransmit. This monitoring node performs as centralized node as well variation amid buffer length and number of packets being sent can be predicted. This information helps in differentiating spurious retransmission from the packet loss.Design/methodology/approachBased on transmission detection, action is accomplished whether to retransmit or evade transmission. Monitoring node selection is achieved by presenting improved cuckoo search algorithm. The modified support vector machine algorithm is greatly used for variation-based spurious transmission.FindingsThe research work which is attained by ST-EPRM. The packet reordering necessity is evaded by presenting random linear network coding process on wireless network physical layer which function on basis of sequence numbers. The spurious retransmission happening over wireless network is obtained by presenting monitoring concept for reducing number of spurious retransmissions because it might need more than three DUPACKs for triggering fast retransmit. This monitoring node performs as centralized node as well variation amid buffer length and number of packets being sent can be predicted. This information helps in differentiating spurious retransmission from the packet loss.Originality/valueBased on transmission detection, action is accomplished whether to retransmit or evade transmission. Monitoring node selection is achieved by presenting improved cuckoo search algorithm. The modified support vector machine algorithm is greatly used for variation-based spurious transmission.

Refined Node Energy Consumption Modeling in a LoRaWAN Network.

LPWAN technologies such as LoRa are widely used for the deployment of IoT applications, in particular for use cases requiring wide coverage and low energy consumption. To minimize the maintenance cost, which can become significant when the number of sensors deployed is large, it is essential to optimize the lifetime of nodes, which remains an important research topic. For this reason, it is necessary that it is based on a fine energy consumption model. Unfortunately, many existing consumption models do not take into account the specifications of the LoRaWAN protocol. In this paper, a refined energy consumption model based on in-situ measurements is provided for a LoRaWAN node. This improved model takes into account the number of nodes in the network, the collision probability that depends on the density of sensors, and the number of retransmissions. Results show the influence of the number of nodes in a LoRaWAN network on the energy consumption of a node and demonstrate that the number of sensors that can be integrated into a LoRaWAN network is limited due to the probability of collision.

An SDN‐based technique for reducing handoff times in WiFi networks

SummaryThe handoff process in WiFi networks has a very important role to provide continuity of the service to wireless users where the access points (APs) are in numerous numbers. In traditional WiFi networks, the handoff process takes a few seconds and is initiated by the wireless devices, but these few seconds could be a reason for information loss when dealing with delay sensitive applications. In this paper, we propose a reduced detection and discovery time approach (DeRy), which generates a solution based on software defined WiFi networks (SD‐WiFi). A simple network management protocol (SNMP) manager and received signal strength indicator (RSSI) manager are employed to report the AP traffic conditions and RSSI values to the SDN controller and instead of the wireless devices, the handoff decisions are taken at the controller end. When to initiate the handoff (detection phase) and which destination AP to associate to (discovery phase) are handled by the centralized SDN controller. To implement DeRy, extensive simulation runs are carried out on Mininet‐NS3‐WiFi network simulator. The simulation results show that the DeRy significantly reduces the handoff times by 60–70% and reduces the average number of retransmission by 4–49%, hence maintaining the throughput for delay sensitive applications such as VoIP, when compared to the standard RSSI‐based handoff scheme, channel measurement based access selection scheme (CMAS) and (DL‐SINR) downlink‐signal to interference plus noise ratio AP selection scheme (DASA).

Data Dissemination Scheme for VANET using Genetic algorithm and Particle Swarm Optimization

A vehicular ad hoc network (VANET) consist of moving vehicles connected via wireless technology e.g., Wireless Access in Vehicular Environment (WAVE) for the aim of exchanging information. Therefore data dissemination in VANET has become issue of debate for researcher. In VANET broadcasting play an important role. The aim of VANET is to ensure passenger safety through emergency message. With multiple objectives broadcast storm is assumed to be an NP-Hard problem. In this paper we propose DDV algorithm to solve broadcast storm problem. Fitness function has used to optimize the objective of proposed algorithm. The proposed algorithm producing better optimization results. We are considering a highway scenario in city with dynamic rotation, to evaluate the performance of the DDV algorithm we compare the result with Smart flooding techniques, MOGA (Multi Objective Genetic algorithm) [1] and EEADP. Our result show the better performance in terms of reduce the number of retransmission, increase the packet delivery ratio and provide better throughput.

An Efficient Cluster-based Flooding Using Fuzzy Logic Scheme (CBF2S) In Mobile Adhoc Networks

Aims and Objectives: Mobile Ad hoc Networks (MANET) hold a set of numerous mobile computing devices useful for communication with one another with no centralized control. Due to the inherent features of MANET such as dynamic topology, constrained on bandwidth, energy and computing resources, there is a need to design the routing protocols efficiently. Flooding is a directive for managing traffic since it makes use of only chosen nodes for transmitting data from one node to another. This paper intends to develop a new Cluster-Based Flooding using the Fuzzy Logic Scheme (CBF2S). Methods: To construct clusters and choose proper cluster heads (CHs), the fuzzy logic approach is applied with the use of three parameters namely link quality, node mobility, and node degree. The presented model considerably minimizes the number of retransmissions in the network. Results: The presented model instructs the cluster members (CM) to flood the packets inside a cluster which is known as intra-cluster flooding and CHs to flood the packets among the clusters which is known as inter-cluster flooding. In addition, the gateway sends a packet to another gateway for minimizing unwanted data retransmissions when it comes under different CH. The presented CBF2S is simulated using the NS2 tool under the presence of varying hop count. Conclusion: The CBF2S model exhibits maximum results over the other methods in terms of overhead, communication overhead, traffic load, packet delivery ratio, and the end to end delay.

Learning automata based energy efficient and reliable data delivery routing mechanism in wireless sensor networks

Wireless Sensor Networks (WSNs) emerged as major data gathering paradigm due to its wide variety of applications. Achieving energy efficient reliable data delivery is an important concern in WSNs. In a network, packet loss reduces the reliable data transmission rate. Retransmissions are required to increase the reliable data transmission rate. However, number of retransmissions also increase the network energy consumptions and packet delivery delay. Packet loss rate can be reduced by selecting a best successor node in a routing path. In this work, a Learning Automata based Routing mechanism is proposed for wireless sensor networks to achieve energy efficiency and reliable data delivery. In this paper, learning automata has been adopted to calculate next node’s selection probability in a routing path using node’s score, quality of link and previous selection probability. Further, an energy efficient reliable routing mechanism is proposed using combination of learning automata and A-star search algorithm. The proposed mechanism determines energy efficient and reliable routing paths in WSNs by favoring highest residual energy, good quality of link , more free buffer space and less distance. Finally, simulation results indicate that the proposed algorithm reduces the energy consumptions, data delivery delay, data transmissions and increases the network lifetime.

A Distributed Congestion Control Strategy Using Harmonic Search Algorithm in Internet of Vehicles

Aiming at the diversified requirements of network application QoS (Quality of Service) in the terminal equipment of Internet of Vehicles, this paper proposes a distributed congestion control strategy based on harmony search algorithm and the Throughput Evaluation Priority Adjustment Model (TEPAM) to ensure real-time transmission of high-priority data messages related to security applications. Firstly, the channel usage rate is periodically detected and the congestion is judged; then, in order to minimize delay and delay jitter as the goal, harmony search algorithm is utilized to perform global search to obtain a better solution for the transmission range and transmission rate. Secondly, packet priority and the TEPAM are applied to indicate the sending right of each packet. The data message priority and throughput percentage factor are used to express the transmission weight of each data message. Besides, the real-time evaluation of path state in MPTCP is carried out by the batch estimation theory model, which realizes the on-demand dynamic adjustment of the network congestion time window. Finally, SUMO, MOVE, and NS2 tools are used to create a VANET-like environment to evaluate the performance of the proposed congestion control strategy. Experimental results show that the proposed method is superior to other three methods in the four indicators of average delay time, average transmission rate, number of retransmissions, and packet loss rate compared with other advanced methods.

ПРОГНОЗИРОВАНИЕ ВРЕМЕНИ РАБОТЫ БЕСПРОВОДНЫХ СЕНСОРНЫХ СЕТЕЙ НА ОСНОВЕ ТЕХНОЛОГИИ LORA С ИСПОЛЬЗОВАНИЕМ ВЕРОЯТНОСТНОЙ МОДЕЛИ МАРКОВСКИХ ЦЕПЕЙ

Real-time monitoring systems and sensors are not complete without wireless data transmission modules. Improving energy efficiency requires examining various system parameters that affect the power consumption of transmitting and receiving devices. The most important parameter of any autonomous wireless network is its uptime. In this work, we used LoRa wireless modules on the SX1278 chip manufactured by Semtech to determine their power consumption in various operating modes. The obtained data were used to build a consumption model of the device when connected to a receiver. Three operating modes are considered: transmit mode, receive mode and sleep mode. In an ideal communication channel, all transmitted data reaches the receiver with 100% probability. In a real situation, data reaches the receiver with a certain probability, depending on the communication channel, transmission power, distance to the addressee, and network parameters. In this work, the occurrence of an error is random. In this case, the occurrence of an error during reception entails a lack of confirmation of receipt or a request for re-sending data. Sending data again increases the power consumption of the device and, consequently, decreases the operating time of the wireless device. This paper shows the dependences of the operating time on various initial monitored parameters of the device, such as: confirmation timeout, packet length, time of one transmission cycle and the maximum number of retransmissions in one cycle. The developed model for predicting the consumption of the device can be used in the design of autonomous wireless sensor monitoring networks.

WITHDRAWN: Effective adaptive routing for Lossy networks using enhanced RPL in the heterogeneous network

Abstract In this network environment, the different applications of sensor have to generate a huge data traffic which includes basic sensing methods such as air pollution or environment checking. In this huge congested network, the two-standard objective methods, that are objective method 0 (OM0) and objective method 1 OM1 for Congested Networks that has to perform deprived routing paths by picking an already congested parental node and a greater number of retransmissions happened between the networks. So, it is required that to design a new effective Routing Protocol for various congested environments. In our paper, I observe the RPL protocol for various congested networks and developed a new routing protocol technique (QWL-RPL) related with information of queue and node-wise workload. Objective of our developed Routing protocol has to attain an Optimum shortest path in challengeable performance. Our proposed Objective method has to consider connectivity capacity and also representing status of congestion in particular node or device using information of queue in data or packet. Our proposed routing technique was implemented in Cooja environment for comparing with the existing techniques. The results of simulation are shown through Queue status information and balanced Workload routing protocol of RPL that can reduce the data traffic of a various networks as related with OM0 and OM1, exactly specified terms of average delay, ratio of packets transmission and reception, jitter delay and header overhead information. At last, the results shown as on average, 30%–50% reduced header overhead information, 15%–30% reduced in average throughput delay, 10%–30% improvement in PRR and 25%–40% reduced in jitter delay.

Hierarchical Scheduling for Suppression of Fronthaul Delay in C-RAN with Dynamic Functional Split

The cloud radio access network (C-RAN) can potentially reduce a network's deployment cost and energy consumption. However, a connection between a centralized baseband unit and distributed remote radio heads, known as a fronthaul, introduces an additional delay to both control and user planes. This delay is a serious limiting factor for radio resource management functionalities, such as scheduling, because the radio resources are assigned to users according to outdated channel quality information if the fronthaul delay is non-zero. This article provides an overview of existing scheduling approaches suitable for C-RAN and identifies their potential limitations. Based on these limitations, we outline a framework for hierarchical scheduling. The hierarchical scheduling mitigates a negative impact of the fronthaul delay on the throughput of non-cell-edge users and enables efficient retransmission of erroneous data. Besides, cell-edge users can still benefit from interference mitigation techniques requiring centralized control. We compare individual scheduling approaches and show that hierarchical scheduling increases the network throughput (by up to 26 percent) and reduces the number of retransmissions with respect to the existing solutions.

Estimating and predicting link quality in wireless IoT networks

The use of poor-quality links in Internet of Things (IoT) networks leads to a bad quality of experience (QoE) with long delivery delays, low reliability, short lifetime of battery-operated nodes, to name but a few. In addition, network resources, such as bandwidth and node energy, are wasted by retransmissions. An accurate estimation of link quality will enable the network to better select the links used for data gathering. Hence, the number of retransmissions needed to achieve the required end-to-end reliability is decreased, leading to shorter end-to-end delivery times, a higher network throughput and an increased network lifetime. In this paper, different link quality estimators are reviewed with a more particular focus on those based on Received Signal Strength Indicator (RSSI) and Packet Delivery Ratio (PDR). We propose to go even further than link quality estimation with link quality prediction. The expected benefit of using link quality prediction is to anticipate link breakages and route changes before loosing packets. It should result in a better QoE provided by the network. For that purpose, we evaluate the performance of four machine learning techniques (i.e. Linear Support Vector Machine, Logistic Regression, Support Vector Machine and Random Forest) working on the traces collected from a real IoT network. They are compared in terms of per-class metrics as well as global metrics. In addition, issues dealing with the deployment of such machine learning techniques in IoT networks with limited resources and energy are presented.