You are currently viewing a new version of our website. To view the old version click .
MDPIMDPI
Search all articles
Sensors
Download PDF
  • Article
  • Open Access

4 February 2013

A Routing Protocol Based on Energy and Link Quality for Internet of Things Applications

,
,
,
,
and
1
Federal University of Minas Gerais, Av. Antonio Carlos, 6627, CEP 31270-901, Belo Horizonte, Brazil
2
Federal University of Pará, Av. Augusto Corrêa, 01, CEP 66075-110, Belém, Brazil
3
Institute of Computer Science and Applied Mathematics, University of Bern, Neubrückstrasse 10,3012, Bern, Switzerland
4
Federal University of Ceará, Campus do Pici, Bloco 942-A, CEP 60020-181, Fortaleza, Brazil
Sensors2013, 13(2), 1942-1964;https://doi.org/10.3390/s130201942
This article belongs to the Special Issue Ubiquitous Sensing
Version Notes
Order Reprints

Abstract

The Internet of Things (IoT) is attracting considerable attention from the universities, industries, citizens and governments for applications, such as healthcare,environmental monitoring and smart buildings. IoT enables network connectivity between smart devices at all times, everywhere, and about everything. In this context, Wireless Sensor Networks (WSNs) play an important role in increasing the ubiquity of networks with smart devices that are low-cost and easy to deploy. However, sensor nodes are restricted in terms of energy, processing and memory. Additionally, low-power radios are very sensitive to noise, interference and multipath distortions. In this context, this article proposes a routing protocol based on Routing by Energy and Link quality (REL) for IoT applications. To increase reliability and energy-efficiency, REL selects routes on the basis of a proposed end-to-end link quality estimator mechanism, residual energy and hop count. Furthermore, REL proposes an event-driven mechanism to provide load balancing and avoid the premature energy depletion of nodes/networks. Performance evaluations were carried out using simulation and testbed experiments to show the impact and benefits of REL in small and large-scale networks. The results show that REL increases the network lifetime and services availability, as well as the quality of service of IoT applications. It also provides an even distribution of scarce network resources and reduces the packet loss rate, compared with the performance of well-known protocols.

1. Introduction

The popularization of wireless and sensor technologies, allied to the demand for new Internet of Things (IoT) applications, is creating a new ubiquitous and smart IoT applications era. The IoT [1] is composed of a set of technologies that provide connectivity at all times, everywhere, and about everything. IoT is based on the principle that objects or things interact and cooperate with each other by using wireless links to ensure ubiquitous communications. They might be Radio-Frequency IDentification (RFID) tags, sensor nodes, actuators, or mobile phones, among others. In this context,Wireless Sensor Networks (WSNs) [2,3] play an important role in providing ubiquitous computing that's capable of connecting both real and virtual worlds.

WSNs/IoT applications have a great impact on the quality of life of people and also lead to economic benefits. Thus, IoT/WSNs are attracting considerable attention from universities, industries and governments in assisting the development of new technologies and applications, such as comfortable homes and offices, healthcare, environmental monitoring and smart cities. For example, the ubiquitous systems and wireless sensor technologies offer suitable solution for improving the efficiency of the food supply chain [4,5]. In the case of applications for healthcare, patients can carry medical sensors to monitor key parameters, such as body temperature, blood pressure, ECG (electrocardiogram) and breathing. Furthermore, medical centers will be able to perform advanced remote monitoring to assess patients condition [6]. Regarding real implementations, the Smart Santana project [7] proposes an experimental research facility in a city to support typical applications and services for smart cities [8]. The facility comprises more than 20,000 IoT devices, divided into topologies that have tens or hundreds of nodes, depending on the applications, such as environmental monitoring and smart parking.

In many IoT applications, the sensed data must be sent to the Base Station (BS) for further operations. This should be accomplished through efficient routing protocols that are key components to improve the data transmission, energy-efficiency, and scalability in WSNs. However, the characteristics of WSNs/IoT raise many challenges in designing efficient communication protocols, owing to limited resources and the unreliability of low-power wireless links that typically lack in terms of Quality of Service (QoS)requirements. At the same time, there still remains a need to find a multipath-aware routing protocol that assures data transmission with low delay, latency, loss rate and minimum energy consumption for various IoT applications [9].

In the context of routing protocols, one important criterion used in the route discovery process is the quality estimation of the communication links between nodes. This quality is usually measured as a single value, such as Received Signal Strength Indicator (RSSI) or Link Quality Indicator (LQI) [10]. However, LQI/RSSI only represents a snapshot at a specific point in time for one link between two nodes, and lacks any additional information about remaining energy, hop count and end-to-end. Thus, there is still an urgent need to find a reliable scheme to estimate the end-to-end link quality based on information of different layers (cross-layer) [11], and it is important to enable the nodes that can transmit packets with a high degree of reliability, while increasing the network lifetime and the overall QoS support for IoT applications.

According to Rocha et al. [12], as a result of recent technological advances in embedded systems,the processing and memory constraints in WSNs are tending to disappear. However, the problem of energy constraint remains a critical issue. Furthermore, an external energy supply is usually unavailable and the replacement of batteries is not feasible for large-scale networks as expected for IoT systems. Thus, two of the main design objectives of WSN/IoT applications are to reduce energy consumption and prolong the network lifetime.

Another important issue for WSNs/IoT applications is how to mitigate the energy hole or hot-spot problem [13], where the nodes located closest to the BS or in the most used paths tend to use up their energy resources prematurely [14]. This leads to a loss of connectivity between a node near to BS and the packets that are sent but not received at BS, which causes a wasteful of energy and wireless resources. Thus, the route selection scheme must consider the residual energy and end-to-end link quality to avoid the energy holes, while at the same time providing load balance and fair distribution of the scarce network resources.

Given the characteristics outlined above, a routing protocol for WSN/IoT scenarios must also minimize the signaling overhead, which is responsible for increasing the processing and energy consumption in the nodes as well as increasing the packet loss and delay in the network. Additionally,reliable routes must be selected by using a mechanism to estimate the end-to-end link quality, base don cross-layer information, such as network conditions, energy and hop count. However, the current routing protocols for IoT applications do not take into account these key features, and novel solutions must be created.

This article addresses the problems of ensuring reliability, together with energy-efficiency and load balancing in flat-based (homogeneous nodes) WSN/IoT architectures, by proposing an extended version of a routing protocol based on energy and link quality information (REL) [15]. REL aims to overcome the drawbacks that have been discussed earlier and allows the data transmission with low latency,packet loss, and high reliability, as well as a fair distribution of wireless resources, while increasing the network lifetime, for various flat-based IoT applications, such as smart parking, intrusion detection,and monitoring of river flows.

REL proposes an end-to-end route selection scheme based on cross-layer information with a minimum overhead. To achieve energy-efficiency, the nodes send their residual energy to neighboring nodes by means of an on-demand piggyback scheme. Additionally, REL uses an event-driven mechanism to provide load balancing and avoid the energy hole problem.

This article extends our previous work [15] as follows: (i) it optimizes the process of route selection by means of end-to-end link quality evaluation and energy information; (ii) it employs a new method for smoothing the link quality estimation; and (iii) it conducts an in-depth analysis of the proposed solution for large-scale flat-based networks. It presents new simulation and testbed experiments to show the effects and benefits of REL for IoT systems. The results show that, compared with the performance of other well-known routing protocols, REL increases the network lifetime, enhances service availability,and ensures QoS support for IoT applications, while reducing the packet loss rate and signaling overhead.

The remainder of this paper is structured as follows. Section 2 outlines the existing routing protocol sand their drawbacks compared with REL. Section 3 describes the proposed routing by the energy and link quality (REL) protocol for IoT applications. Section 4 shows simulation and testbed experiments. Section 5 summarizes the main contributions and results of this article.

4. Performance Evaluation

REL is evaluated through a testbed (small-scale) and OMNET++ simulator (large-scale). The testbed results are used to calibrate the simulation experiments, and also to analyze the performance of REL in real experiments. The objective of the simulation experiments is to evaluate the REL and compare it with both AODV and LABILE in terms of energy-efficiency, latency and data delivery in a large-scale scenario.

4.1. Parameter Calibration

Testbed experiments were conducted to find LQIth and Eth, since they can be used to evaluate the link quality in real-world deployments in the presence of interference and moving people, as expected for IoT scenarios. Additionally, in testbed experiments, the energy consumption can be calculated with real hardware used by many networks.

The experiments were carried out in the 2nd floor of the Computer Engineering Department at the Federal University of Pará. The experiments consisted of 10 SunSPOT motes [31] in a grid topology of3 × 3 and the BS on the edge, as depicted in Figure 2. The proposed grid topology is able to provide a minimum number of 5 neighboring nodes for each node.

Figure 2. Testbed topology and connectivity map.

The experiments were carried out in an indoor environment, as shown in Figure 3. In this scenario, the nodes were deployed on the roof of the rooms and in the corridor. The experiments were conducted in a real world environment in the presence of continuous movement of people and other wireless network devices, which caused noise and interference.

Figure 3. Testbed scenario on Computer Engineering Department

As mentioned earlier, most of the IoT applications require reliable communications systems to allow the end-user to analyze the accurate collected data by the sensor nodes. We assume that most of the IoT applications need a PDR that is higher than 80% as QoS requirements (this value can be easily changed for different applications). In this context, experiments were conducted to find the LQI threshold for this purpose. Based on the results shown in Figure 4, we selected LQIth= 220.

Figure 4. Testbed experiment to find LQIth.

Experiments were also conducted to find the Eth in the real hardware since it is possible to analyze energy consumption in a real device. Figure 5 shows five possible values for the energy threshold varying from 2 to 10. Values of Eth higher than 6 reduce the network lifetime. This occurs because of the low frequency of the RADV notifications, which provides an unbalanced consumption of network resources,degrades the performance of the load balancing mechanism, and the use of multiple routes. As result,nodes located at potential energy holes take a longer time to be detected. On the other hand, Eth= 2 has the highest lifetime of all, exceeding the value 4 by 2 min.

Figure 5. Testbed experiment to find Eth.

4.2. Testbed Experiments

Testbed experiments were also conducted to evaluate REL and compare it with AODV in terms of network lifetime and data delivery in experiments into real environment, as expected for IoT and ubiquitous computing applications. The experiments were set up to allow each node to send 1200 packets to the BS with an inter packet interval of 0.5 s. This configuration is common for most IoT applications that require a high data rate.

Figure 6 shows the results of PDR for each node when they use REL or AODV as the routing protocol. It important to point out that although PDR ranges from 0% to 100%, Figure 6 establishes an interval from 90% to 100% to make the performance evaluation clearer. On the basis of the results of Figure 6,the total packet loss of AODV is 67.6% and REL is 28.3%. Thus, REL increases the PDR by 39.3%. This is due to the fact that REL selects routes on the basis of cross-layer information, with end-to-endlink quality estimation (QoS-aware), unlike AODV that only chooses paths based on hop count. Thus,REL is able to provide data transmission with a higher degree of reliability in IoT scenarios.

Figure 6. Packet delivery ratio in testbed experiments.

The PDR of node 5 has a higher value (13.3%) of packet loss using AODV as the routing protocol,and only 1.1% using REL. As depicted Figure 2, node 5 is the central node of the grid topology used forth experiments. REL selects routes together with load balance operation, and, thus avoids continually sending packets to the same node, which causes interference, collisions, and buffer overflow. On the other hand, AODV selects routes that are only based on the minimum number of hops, therefore nodes7, 8 and 9 often send packets to node 5.

The Figure 7 shows the energy-efficiency for this testbed. AODV presents unbalanced energy consumption between nodes and their possible premature death, as can be seen in Figure 7(a). Additionally, five nodes consumed up to 50% of the energy resources before the end of the experiment. As expected, these nodes are the same as those with the highest packet loss values. Thus, routes less reliable and more susceptible to errors require more retransmissions and spend more energy.

Figure 7. Network lifetime in testbed experiments, (a) AODV, (b) REL.

In contrast, REL increases the network lifetime by more than 15% compared with AODV and also presents an almost uniform distribution of energy consumption, as shown in Figure 7(b). This is due to the fact that REL has a mechanism for opportunistic piggybacking, which leads to an overhead reduction and, thus, improves the energy consumption. Additionally, REL uses a load balancing scheme that provided uniform energy consumption for all network nodes.

Finally, the difference between the higher and lower energy consumption for REL was 6% (nodes 1 and 2), whereas for AODV this value exceeded 23% (nodes 3, 4, 6, 8, and 9). This is due to the reasons explained above.

4.3. Simulation

Simulation experiments were conducted to analyze the performance of REL using the Castalia Framework [32], which is a widely used network simulator for WSNs based on OMNET++simulator [33]. The simulations were carried out and repeated 20 times with different random seed numbers in order to obtain a confidence interval of 95%.

Due to limited available nodes in the testbed, simulation experiments were conducted to evaluate the REL and compare its performance with AODV and LABILE in terms of energy-efficiency, latency and data delivery in a large-scale scenario, i.e., using a monitoring area of 100 m × 100 m and upto 200 nodes. AODV is a standard reference, where can be considered as a benchmark solution for flat routing protocols and scenarios. The LABILE protocol presents an assessment strategy based onend-to-end link quality, thus making it suitable and essential for assessing and comparing its performance with REL.

The values for energy and LQI thresholds in the simulations were collected in the testbed experiments. Table 1 shows the threshold values and other simulation parameters. The topology was generated by using Bonn Motion [34] tool, where nodes are uniformly distributed in the simulation.

Table 1. Simulation parameters.

Similar to LQI and energy thresholds, REL has also a hop count threshold, which cannot be found in the testbed experiments due to of the limited number of nodes. Thus, simulation experiments were performed to find HCdiffmax_allow, which is able to provide a reasonable number of alternate routes without affecting the latency and PDR.

Figure 8 shows PDR for different values of HCdiffmax_allow, which ranges from 2 to 10 hops. All possible values have a PDR higher than desired, i.e., 80%. There is a need to choose a suitable value for LQIth, which creates a large number of possible routes to be used for the load balancing scheme. However, values higher than 6 hops create more routes and the signaling overhead can have an effect on the network lifetime and latency. On the basis of our results, the most suitable values for HCdiffmax_alloware 2 and 4 hops. In the case of our simulation experiments, we chose HCdiffmax_allow of 4 hops to avoid a small number of alternate routes. The value 4 is able to consider more routes and achieve a better PDRresult, as shown in Figure 8.

Figure 8. Simulation results of HCdiffmax_allow.

Starting with the scenario of Table 1, other configurations were setup to assess the performance of REL compared with AODV and LABILE. The additional scenarios included configurations that changed the number of nodes to 20, 40, 60, 80, and 200. The purpose of this variation was to have scenarios with different node traffic demands as expected in real IoT systems. During our data analysis, the scenario with 20 nodes was used to check the accuracy of the simulation results, by comparing them with the testbed experiments.

Additionally, the scenarios that have a different number of nodes create alternate routes that are useful to evaluate the load balancing and energy-efficiency mechanisms. The performance of REL was compared with AODV and LABILE in terms of network lifetime, latency and PDR.

The WSNs/IoT applications have typically heterogeneous traffic characteristics, which can be file transfer or sensing data. Given this fact, protocols developed for these scenarios should regard the latency as an important metric. In this context, during our experiments, we evaluated the latency of REL, AODV and LABILE. Each received packet was categorized into one of the three latency categories presented in Table 2. This classification aims to clarify the performance analysis of the evaluated protocols.

Table 2. Latency.

For scenarios with 20 nodes, results show that REL improves the latency up to 0.8%, the latency between 0 and 40 ms, and 0.4% for packets above 80 ms. LABILE obtained results similar to REL for scenarios with 20, 40 and 60 nodes, where the maximum difference between them was 1% in packets with latency between 0 and 40 ms. However, if the number of nodes increases, the difference rises to 4%for scenarios with 100 nodes.

Additionally, the performance of AODV decreases by up to 17.9% compared with REL, and 13.9%compared with LABILE, for latency between 0 and 40 ms, and for scenarios with to 100 nodes. In the case where the scenarios with few numbers of nodes, the improvement of latency result is reduced due to the availability of short routes. However, in large-scale scenarios, such as 80, 100, and 200, nodes achieve good gains when the REL protocol is used.

In a scenario with 80 nodes, REL exceeded LABILE by up 2.4% in the lower delay category and 1%in the highest delay. The performance gain increases in a scenario with 100 nodes, where the differences are 4% and 1.8%, respectively. These results show the efficiency of REL for improving the usage of network resources.

In the scenario with 200 nodes, the protocol AODV reaches more than 40 ms delay in 26.4% of transmitted packets, a result 6% lower than obtained in the scenario of 100 nodes due to the high number of overhead and interference created in the densest scenario. Regarding the protocol LABILE,the difference reaches 17.5% and 21.6% compared with REL. In the category of delays up to 40 ms, REL maintains the highest performance, exceeding in 4.1% and 20.6% LABILE and AODV, respectively.

The PDR was used to measure the reliability of REL, LABILE and AODV. Figure 9 shows the PDR for all packet transmissions in scenarios with different node densities. Figure 9 establishes an interval from 80% to 100% to make it easier to analyze the results.

Figure 9. Packet delivery ratio in simulation experiments. (a) 20 Nodes, (b) 40 Nodes,(c) 60 Nodes, (d) 80 Nodes, (e) 100 Nodes, (f) 200 Nodes.

As can be seen in Figure 9, REL outperforms AODV and LABILE in all experiments. The difference increases when there is a rise in the number of nodes in the simulated scenario. This can be explained by the load balancing scheme and reliable route selection proposed in REL, when compared with LABILE and AODV that do not have load balancing solutions and select routes based only on link quality and the number of hops.

The scenario with 20 nodes confirms the accuracy of the simulation results in comparison with the testbed is presented in Figure 2. In the case of the testbed experiments, the average value of PDR for AODV was 92.4% and for REL was 96.8%. These results reflect the close approximation of the real and simulated experiments.

In a similar way to what was found in the latency results, there is an improvement of REL, in terms of PDR, compared with AODV and LABILE when the node density increases. These scenarios with high node density are expected for many IoT applications, such as automation of homes and offices, smart parking, healthcare and environmental monitoring.

The scenarios with 60 and 80 nodes show results that correspond to the increasing number of nodes. REL improves by 8.8% when compared with AODV and 4.5% with LABILE in a scenario with 60 nodes. In the scenario with 100 nodes, REL increases the number of packets delivered in 9.4% and 7.9%,respectively. This is due to the fact that increasing the node density also increases the number of available routes for load balancing and causes the phenomenon of signal attenuation. In this context, the lack of a technique to smooth the variation of LQI values leads to a greater switching of routes in LABILE,and thus reduces the data delivery. For these reasons, although LABILE has a mechanism to evaluateend-to-end link quality, its performance was below of the REL protocol. On the other hand, AODV selects routes that are only based on minimum hop count, which reduces its overall reliability.

The analysis showed that results with 100 nodes, i.e., those with high density, had a PDR difference that reached 12% when compared with AODV, and 9.2% when compared with LABILE. Thus, if the results of the high PDR and low latency provided by REL are combined, we can observe high degrees of efficiency and reliability in its route selection and load balancing scheme. These results confirm that REL achieves a level of QoS that is appropriate for ubiquitous computing and IoT applications.

The values of the PDR for systems with 200 nodes show that REL keeps its superior performance even in dense topologies, where the PDR is around 94.5%. For the same scenarios, AODV and LABILE results were 80.8% and 83.3%, respectively.

As mentioned earlier, despite recent technological advances in embedded systems, the energy constraint remains a critical factor in WSNs/IoT, given the fact that sensor nodes have limited energy resources. Furthermore, an external energy supply is usually unavailable and the replacement of batteries is not feasible for large-scale WSNs. In this context, it is essential to have an integrated routing and load balancing scheme that can extend the network lifetime without affecting its reliability.

Figure 10 shows the percentage of nodes that remained alive during the simulation. Two metrics were employed to evaluate the energy-efficiency: the entire lifetime of the network and the saturation time. The network lifetime was measured as the time spent by sensor nodes until the point when only 1% of the network nodes remains alive. The saturation time is similar to the network lifetime, although it starts from the moment of the first node death. The saturation time is useful in assessing the capacity of the routing protocols to dynamically adapt to new topologies resulting from the death of the nodes.

Figure 10. Network lifetime in simulation experiments. (a) 20 Nodes, (b) 40 Nodes,(c) 60 Nodes, (d) 80 Nodes, (e) 100 Nodes, (f) 200 Nodes.

The lack of energy management in existing protocols, such as AODV and LABILE, means that their results can be similar in most of the experiments. The experiments outlined in Figure 10 show differences between the percentage of nodes that were alive during the saturation period for these protocols, although the difference in the network lifetime was a maximum of 1 minute. The reason for this is that LABILE is a proposal that extends the AODV protocol, but does not improved energy-efficiency features.

AODV and LABILE do not provide mechanisms that can take advantage of the large number of network nodes with respect to energy. As a result, in the scenario with 60, 80 and 100 nodes(Figure 10(c–e)) these protocols do not exceed a lifetime of 42 min. Furthermore, the saturation time also remains stable.

When the protocol REL is used, the saturation time in these three scenarios with more nodes also increased by up to 8 min. The saturation time reached 32 min for scenarios of 60 and 80 nodes, while the scenario with 100 nodes reached a maximum of 36 min.

Since the beginning of the saturation period occurs at the moment of the death of the first node, it is also included in the assessment of the energy-efficiency of a protocol. The maximum time spent until the first death, when the protocols AODV and LABILE were used, was 12 min. For the REL protocol, the minimum value was 18 min in the scenario with 20 nodes, and 22 min in topologies with more nodes.

The beginning of saturation occurs earlier in the scenario with 200 nodes compared with the scenario with 100. The instant 18 starts to decrease the number of nodes alive and extends up to the instant56. The first death occurs firstly in the scenario with 200 nodes due to interference generated by the large number of nodes. However, the saturation time is high due to the large number of alternate routes,a feature that favors the load balancing and adaptability of the REL. Protocols AODV and LABILE showed a reduction in total network lifetime, decreasing by up to 8 min.

5. Conclusion and Future Works

This article has presented a routing protocol together with load balance scheme based on energy and link quality (REL) for IoT applications, such as automation of comfortable homes and offices, healthcare,environmental monitoring, and smart parking. REL combines a reliable scheme for route discovery and load balance mechanism, which provides high reliability, QoS-awareness and energy-efficiency Moreover, it proposes an end-to-end route selection scheme based on cross-layer information with a minimal overhead. Nodes become energy efficient by sending the residual energy to their neighboring nodes with the aid of a piggyback and on-demand scheme. Additionally, REL also uses an event-drivenmechanism to provide load balancing as a way to improve the system performance and avoid the energy hole problem.

REL was evaluated by using simulation and testbed experiments to show its effects and benefits when compared with existing solutions. Testbed results were used to calibrate and confirm the accuracy of the simulation experiments, as well as present the impact of REL in real scenarios. At the sametime, simulation experiments were useful to evaluate REL and compare it with well-known AODV and LABILE protocols, in terms of energy-efficiency, latency and data delivery in large-scale scenarios as expected for many IoT applications. In large-scale networks with a high node density, the results showed that REL increases the network lifetime by up to 26.6%, latency by 17.9% and packet delivery by 12%,compared with AODV and LABILE.

As future studies, we will implement and evaluate REL in a large-scale test-bed, such as Wise bed and Smart Santander. We will also manage the radios/transmissions to improve the energy saving process.

Acknowledgments

This work was supported by the National Council for Scientific and Technological Development(CNPq), CTIC RNP CIA2, and the Foundation for Research of the State of Pará (FAPESPA), Brazil.

References

  1. Atzori, L.; Iera, A.; Morabito, G. The internet of things: A survey. Comput. Netw. 2010, 54, 2787–2805. [Google Scholar]
  2. Akyildiz, I.; Su, W.; Sankarasubramaniam, Y.; Cayirci, E. Wireless sensor networks: A survey. Comput. Netw. 2002, 38, 393–422. [Google Scholar]
  3. Yick, J.; Mukherjee, B.; Ghosal, D. Wireless sensor network survey. Comput. Netw. 2008, 52, 2292–2330. [Google Scholar]
  4. Dada, A.; Thiesse, F. Sensor applications in the supply chain: The example of quality-based issuing of perishables. The Internet of Things 2008, 4592, 140–154. [Google Scholar]
  5. Ilic, A.; Staake, T.; Fleisch, E. Using sensor information to reduce the carbon footprint of perishable goods. IEEE Pervasive Comput. 2009, 8, 22–29. [Google Scholar]
  6. Gao, T.; Greenspan, D.; Welsh, M.; Juang, R.; Alm, A. Vital Signs Monitoring and Patient Tracking over a Wireless Network. Proceedings of the 27th Annual International Conference of the Engineering in Medicine and Biology Society, Osaka, Japan, 30 August–3 September 2006; pp. 102–105.
  7. Chatzigiannakis, I.; Mylonas, G.; Vitaletti, A. Urban pervasive applications: Challenges, scenarios and case studies. Comput. Sci. Rev. 2011, 5, 103–118. [Google Scholar]
  8. Sanchez, L.; Galache, J.; Gutierrez, V.; Hernandez, J.; Bernat, J.; Gluhak, A.; Garcia, T. Smart Santander: The Meeting Point between Future Internet Research and Experimentation and the Smart Cities. Proceedings of the 2011 IEEE Future Network & Mobile Summit, Warsaw,Poland, 15– 17 June 2011; pp. 1–8.
  9. Radi, M.; Dezfouli, B.; Bakar, K.; Lee, M. Multipath routing in wireless sensor networks: Surveyand research challenges. Sensors 2012, 12, 650–685. [Google Scholar]
  10. Renner, C.; Ernst, S.; Weyer, C.; Turau, V. Prediction Accuracy of Link-Quality Estimators. Proceedings of the 8th European Conference on Wireless Sensor Networks, Bonn, Germany, 23–25 February 2011; pp. 1–16.
  11. Gomez, C.; Boix, A.; Paradells, J. Impact of LQI-based routing metrics on the performance of aone-to-one routing protocol for IEEE 802.15.4 multihop networks. EURASIP J. Wireless Commun. Netw. 2010, 2010, 1–20. [Google Scholar]
  12. Rocha, A.; Pirmez, L.; Delicato, F.; Lemos, É; Santos, I.; Gomes, D.; de Souza, J. WSNs clustering based on semantic neighborhood relationships. Comput. Netw. 2012, 5, 1627–1645. [Google Scholar]
  13. Ramos, H.S.; Oliveira, E.M.R.; Boukerche, A.; Loureiro, A.A. Characterization and Mitigation of the Energy Hole Problem of Many-to-One Communication in Wireless Sensor Networks. Proceeding of the IEEE International Conference on Computing, Networking and Communications, Okinawa, Japan, 5– 7 December 2012.
  14. Ramos, H.S.; Guidoni, D.; Boukerche, A.; Nakamura, E.F.; Frery, A.C.; Loureiro, A.A. Topology-Related Modeling and Characterization of Wireless Sensor Networks. Proceedings of the 8th ACM Symposium on Performance Evaluation of Wireless Ad Hoc, Sensor, and Ubiquitous Networks, Miami Beach, FL, USA, 31 October–4 November 2011; p. 33.
  15. Machado, K.; Rosário, D.; Nakamura, E.; Abelém, A.; Cerqueira, E. Design of a Routing Protocol Using Remaining Energy and Link Quality Indicator (REL). Proceedings of the 6th Latin America Networking Conference, Quito, Ecuador, 12–14 October 2011; pp. 33–39.
  16. Ehsan, S.; Hamdaoui, B. A survey on energy-efficient routing techniques with QoS assurances for wireless multimedia sensor networks. IEEE Commun. Surv. Tutor. 2012, 14, 265–278. [Google Scholar]
  17. Becker, M.; Gupta, A.; Marot, M.; Singh, H. Improving Clustering Techniques in Wireless Sensor Networks Using Thinning Process. Proceedings of the International Conference on Performance Evaluation of Computer and Communication Systems: Milestones and Future Challenges, Lisbon,Portugal, December 2011; pp. 203–214.
  18. Rosário, D.; Costa, R.M.; Paraense, H.; Machado, K.; Cerqueira, E.; Braun, T. A Smart Multi-Hop Hierarchical Routing Protocol for Efficient Video Communication over Wireless Multimedia Sensor Network. Proceedings of the 2nd IEEE International Workshop on Smart Communication Protocols and Algorithms, Ottawa, Canada, 10– 15 June 2012.
  19. Rosário, D.; Machado, K.; Abelém, A.; Monteiro, D.; Cerqueira, E. Recent Advances and Challenges in Wireless Multimedia Sensor Networks; Mobile Multimedia—User and Technology Perspectives; In Tech: New York, NY, USA, 2012; pp. 74–96. [Google Scholar]
  20. Diallo, C.; Marot, M.; Becker, M. A Distributed Link Quality Based D-Clustering Protocol for Dense Zig Bee Sensor Networks. Proceedings of the IFIP Wireless Days 2010, Venice, Italy, 20–22 October 2010; pp. 1–6.
  21. Gnawali, O.; Fonseca, R.; Jamieson, K.; Moss, D.; Levis, P. Collection Tree Protocol. Proceedings of the 7th ACM Conference on Embedded Networked Sensor Systems, Berkeley,CA, USA, 4–6 November 2009; pp. 1–14.
  22. Diallo, C.; Marot, M.; Becker, M. Link Quality and Local Load Balancing Routing Mechanisms in Wireless Sensor Networks. Proceedings of the Sixth Advanced International Conference on Telecommunications (AICT), Barcelona, Spain, 9–15 May 2010; pp. 306–315.
  23. Perkins, C.; Belding-Royer, E.; Das, S. Ad hoc on Demand Distance Vector (AODV) Routing (RFC3561). Available online: http://www.ietf.org/rfc/rfc3561.txt (accessed on 30 January 2013).
  24. Butt, M.; Javed, M.; Akbar, A.; Taj, Q.; Lim, C.; Kim, K. Labile: Link Quality-Based Lexical Routing Metric for Reactive Routing Protocols in IEEE 802.15.4 Networks. Proceedings of the 5th International Conference on Future Information Technology (Future Tech), Busan, Korea, 21-23 May 2010; pp. 1–6.
  25. IEEE -TG15.4. 802.15.4: Wireless Medium Access Control (MAC) and Physical Layer(PHY) Specifications for Low-Rate Wireless Personal Area Networks (LR-WPANs); IEEE Standard No.802.15.4-2006; IEEE Standard for Information Technology: Piscataway, NJ, USA, 2006.
  26. Chung, Y. An energy-efficient unicast routing protocol for wireless sensor Networks. Tech. Int. J.Comput. Sci. Emerg. Tech. 2011, 2, 60–64. [Google Scholar]
  27. Li, X.; Lu, J.; Yang, J.; An, J. A Novel Routing and Data Transmission Method for Stub Network of Internet of Things Based on Percolation. Proceedings of the IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), Shanghai, China, 15–19 April 2011; pp. 201–205.
  28. Baccour, N.; Koubaa, A.; Youssef, H.; Jamaa, M.B.; Rosario, D.; Becker, L.B.; Alves, M. F-LQE:A Fuzzy Link Quality Estimator for Wireless Sensor Networks. Proceedings of the 7th European Conference on Wireless Sensor Networks (EWSN), Coimbra, Portugal, 17–19 February 2010; pp. 240–255.
  29. Baccour, N.; Koubaa, A.; Jamaa, M.; Rosario, D.; Youssef, H.; Alves, M.; Becker, L. Radiale: A framework for designing and assessing link quality estimators. Ad Hoc Netw. 2011, 9, 1165–1185. [Google Scholar]
  30. Srinivasan, K.; Dutta, P.; Tavakoli, A.; Levis, P. An empirical study of low-power wireless. ACMTrans. Sens. Netw. 2010, 6, 1–48. [Google Scholar]
  31. Microsystems, S. Sun TM Small Programmable Object Technology (Sun SPOT) Developer's Guide. Available online: http://www.sunspotworld.com/docs/Orange/spot-developers-guide.pdf (accessed on 30 January 2013).
  32. Boulis, A. Castalia, A Simulator for Wireless Sensor Networks and Body Area Networks, Version 2.2. User's Manual; NICTA: Canberra, Australia, 2009. [Google Scholar]
  33. Varga, A. The OMNeT++ Discrete Event Simulation System. Proceedings of the European Simulation Multi conference (ESM 2001), Prague, Czech Republic, 6–9 June 2001.
  34. Aschenbruck, N.; Ernst, R.; Gerhards-Padilla, E.; Schwamborn, M. Bonn Motion: A Mobility Scenario Generation and Analysis Tool. Proceedings of the 3rd International ICST Conference on Simulation Tools and Techniques, Malaga, Spain, 15–19 June 2010; p. 51.

Article Metrics

Citations

Article Access Statistics

Journal Statistics
Multiple requests from the same IP address are counted as one view.
Download PDF
Extended Kalman Filter-Based Methods for Pose Estimation Using Visual, Inertial and Magnetic Sensors: Comparative Analysis and Performance Evaluation
Previous
Numerical Simulation of Optically-Induced Dielectrophoresis Using a Voltage-Transformation-Ratio Model
Next