My First Research on
Cognitive Radio Wireless Sensor Networks
Dated back to Dec 2013
Reza Nourmand
Department of Computer Science and Technology
University of Bedfordshire
Luton, England
Reza.nourmand@studies.beds.ac.uk
Abstract— a cognitive wireless sensor network is the combination of a wireless sensor network and a cognitive radio network enriching wireless sensor network with the opportunistic spectrum access capabilities of cognitive radio networks. In this paper, wireless sensor networks, cognitive radio networks and a combination of both which creates cognitive radio wireless sensor networks is going to be discussed. The strength and weaknesses that this merger created for CR-WSN will be looked at. Research trends are going to be introduced and conducted researches are going to be discussed to some extent. Finally, a list of several open research challenges are going to be given with the aim at drawing the interest of the readers toward the significant issues that should be addressed before the idea of completely independent cognitive radio wireless sensor networks can be understood.
Keywords- wireless sensor network; sensor networks; cognitive sensors; cognitive radio; cognitive radio wireless sensor networks.
I. Introduction
In this paper, we are going to start with introducing wireless sensor networks, move on to the definition of cognitive radio and finally combining these two, creating a new kind of wireless networks called Cognitive Radio Wireless Sensor Networks (CR-WSN). This new kind of network has similar characteristics as wireless sensor networks (WSNs) and it benefits from cognitive radio network’s capabilities regarding dynamic access to the spectrum. This new feature i.e. cognition adds many positive capabilities to WSN and causes some restrictions as well. Afterwards we are going to discuss CR-WSN merits and challenges regarding structure, power consumption, security, etc. Finally, as a sample of the researches being conducted in this area, a number of researches are going to be discussed and research trends alongside with open research areas are going to be introduced. The objective of this paper is to give a rather thorough introduction to the newly introduced cognitive radio wireless sensor network so that the readers can have an overview of the subject, while introducing appropriate literature for further reading if required.
II. Wireless Sensor Networks
According to Vijay et al (2011), wireless sensor networks (WSNs) are networks which include hundreds of small sensors each having processors and memory, spread in the sensing area with short-range wireless communication either between themselves or directly to a node called a sink. A sink node is a node, either fixed or mobile, in charge of gathering sensors’ data and reporting them to users through an existing communication infrastructure or the Internet as can be seen in figure 1. Vijay et al. also point out that WSNs have great influence on military as well as non-military applications including monitoring the weather, security, surveillance, disaster management, battle field applications and in general, any application working with or in need of the data regarding any changes in sound, temperature, light, movement, etc. Their research also shows that reliability and latency are the key factors for the applications of WSNs and should be considered in the implementation of such networks. WSNs are also data-centric networks (Joshi, 2013); hence loss of data or delay would be unacceptable in their applications. They explain many methods to enhance the QoS of WSNs namely Congestion Aware Routing (CAR), MCCP, CODA, etc (Uthra, 2012).
The standard for WSN is IEEE 802.15.4 also known as ZigBee (Cayirci, 2008). WSNs use ISM band to send their data and they experience many collisions because all the free wireless networks are sending their data in this band. So WSNs are both a sufferer from and an interferer for the other wireless networks which use ISM band. This issue has deeply affected the performance of WSNs in terms of power, security, etc.
Figure 1. Conventional Wireless Sensor Networks, Source: Reference (Joshi, 2013)
Therefore, plenty of researches have been conducted around enhancement of WSN performance such as power aware Medium Access Control, development of sensing technique efficiency and cross-layer design methods to resolve this issue. One of these research trends is to pursue Cognitive Radio (CR) technique in WSN.
|
III. cognitive radio
A. Cognitive Radios’ Unique Capabilities
1. Spectrum sensing techniques:
Spectrum sensing is one of the most crucial components of CR which is defined as the ability to analyse, sense, learn, and be aware of parameters associated with the characteristics of radio channel, spectrum and power accessibility, interface and noise temperature, the environment in which radio is operating, applications and the requirements of the end users (Ben Letaief, 2009). In the following section, some techniques of spectrum sensing are going to be introduced briefly based on what Ben Letaief et al. Presented by (Hoven, 2005).
a. Matched Filter Detection: this method of detection is applicable in the environments where the secondary user knows the attributes of primary user’s transmitting signal. This is similar to convolving the unknown signal with a time-reversed version of the template. The major benefit of matched filter is that as a result of coherent detection, it requires less time to reach high processing gain (Zeng, 2007).
b. Energy Detection: this method of detection is applicable when no previous knowledge of primary user’s transmission signal is available to the secondary user. In this case, in order to understand whether the channel is idle or in use, the energy level of the channel would be used as a meter.
c. Wavelet Detection: it is a multi-indicator analysis method where an input signal is broken down into several frequency components, and then each component is analysed with the indicator corresponding to its meter.
d. Co-variance Detection: it is a statistical method based on the co-variance metrics or signal’s autocorrelations. Co-varriance-based signal detection methods are proposed by Gardner in 1992.
e. Cyclostationary Detection: it is a method based on noise uncertainty rather than energy detection. According to Letaief et al. “a signal is said to be cyclostationary (in the wide sense) if its autocorrelation is a periodic function of time t with some period.” (Cabric, 2004).
2. Cooperative Spectrum Sensing
According to Cabric (2004), One of the challenges that CR faces regarding spectrum sensing is the hidden terminal problem. This problem occurs due to either shadowing or multipath fading. As can be seen in figure 2, CR3 is shadowed by the tall building, thus it cannot sense the primary user’s transmission. The method with which this issue can be addressed is collaboration of multiple CRs. According to Ghasemi et al. , cooperation of multiple CRs, significantly raises the probability of detecting the usage of spectrum by the primary user. Basically, cooperative spectrum sensing can be described in the following 3 step process:
Cooperative Spectrum Sensing:
1) All the CRs individually sense the network and decide whether the PU is using the spectrum or not. They keep the record of that decision as a binary variable.
2) Every CR sends the so-called binary variable to the common receiver.
3) The common receiver then makes a decision about whether it can occupy the spectrum or not based on all the binary variables it has received.
Figure 2. Cooperative spectrum sensing in CR networks. CR 1 and CR 3 are shadowed over reporting channel and sensing channel respectively.
Source: (Hoven, 2005)
IV. combining CR with WSN
According to Reference (Cavalcanti, 2008), the communication in WSNs is based on the events. When the expected event occurs, all the sensors which are close enough to sense that event generate a signal to inform the other nodes and their respective sink. Based on the characteristics of these networks, it is expected that a considerable number of nodes would sense each event which might cause a flood of signals generated by these nodes. This flood of signals should be handled in a way that the sink should be able to get informed of the event with sufficient information and that the information should be received by the sink on time, otherwise the information might most likely lose their value or even in some cases it might result into a disaster. Cavalcanti also point out coexistence of several sensor networks in the same area and the fact that they use ISM band as other important reasons that might lead to congestion in such situations. One of the best ways to handle such situations is to combine CR with WSNs. As an advantage of adapting cognitive radio with WSN we can name some capabilities of CR that will be added to WSNs. The following table shows some prospective capabilities of a wireless sensor network with CR (Cavalcanti, 2008).
V. cognitive radio wireless sensor network
As Cavalcanti points out, CR-WSNs are basically a set of geographically scattered self aware energy-constrained sensor nodes each of which has cognitive capabilities. They also mention that the characteristics of these networks are very different than the conventional WSN and conventional CR networks. According to Kaur, wireless sensor networks benefit from several potentials and characteristics of the cognitive radio networks. They mentioned 5 of the most important ones of these so-called potential as followes:
- As opposed to the conventional WSNs, CR-WSNs benefit from dynamic spectrum access. This ability of CR enables WSNs to use a wider range of liscenced as well as unliscenced bands without adding any extra costs. Figure 2 presents a CR-WSNs model making use of ISM band as well as Incumbent band.
- When an event happens, it creates a bursty traffic on the wireless sensor network, but using CR enables WSNs to make a good use of the wasted white spaces in other spectrums to its own benefit in order to handle this traffic.
- One of the most important issues in sensor networks is reducing their power consumption. Power consumption encreases when nodes have to retransmit the packets due to network congestion and the packet loss caused by that. Implementing CR helps WSNs to handle these congestions; hence it leads to resolving the battery consumption issue.
- Making efficient use of multiple concurrent WSN infrastructure and spectrum is another benefits of this combination. As the authers mention, “Dynamic spectrum management may significantly xontribute to the efficient coexistence of spatially overlapping sensor networks in terms of communication performance and resource utilization.”
- Some regulation prohibit networks to use certain bands in certain parts of the world. This problem would not be an issue for WSNs which are using CR.
In the following part of this section, architecture and potential applications of CR-WSNs are going to be discussed considering the changes that CR made to conventional WSNs as Kaur suggests. Then we continue with discussing challenges that CR-WSNs face.
A. CR-WSN Architecture
CR hardly makes any significant changes to WSN architecture rather than facilitating the way nodes communicate between one another and to the sink. CR accomplishes this goal by adding its opportunistic capabilities to WSN regarding spectrum usage and spectrum allocation.
1) CR-WSN Node Structure: With the constraints that sensor networks had left for their successors CR-WSNs regarding power usage, processing performance, memory volume and carrier frequency, CR can only enable nodes in CR-WSN to actively adapt these communication parameters.
2) CR-WSN Topology: Similar to sensor networks, CR-WSNs can have different topologies depending on the particular needs of their application. Regarding their topology, CR-WSNs can be Ad Hoc, Clustered, Heterogeneous and hierarchical, or Mobile.
Figure 3. CR-WSN model, source: Reference (Joshi, 2013)
B. Potential Applications for CR-WSN
Having the above mentioned advantages in comparison with conventional sensor networks, the capabilities of CR-WSNs are applicable in several sensor networks as the ones below.
1) Indoor CR-WSN Applications: These usages include telemedicine, industrial automation, fire and smoke detection and alarm system, etc. Recall that sensor networks use ISM band and these applications are most likely implemented within the environments which are crowded with many devices using this band. Such areas have more crowded and prone to congestion environments; hence using capabilities of CR for sensor networks implemented in these environments would be vital in order to avoid congestion.
2) Multimedia Applications: One of the most important requirements of multimedia is on time delivery of the data. In multimedia if data arrives late it most likely loses its value so congestion and the delay caused by that will not be tolerated. In such applications dynamic spectrum access plays an important role in making the best use of the available unused spectrum in the environment which result in better QoS of Multimedia Usages.
3) Multiclass Heterogeneous Sensing Applications: Some application need to use the data gathered from multiple sensor networks. Congestion between the data signals generated by different sensors in different networks is an issue in such networks. One of the best methods to avoid such congestions is dynamic spectrum management.
4) Real-Time Surveillance Applications: These applications such as target tracking need the least possible delay. This need will not always be met by the conventional WSNs. This problem can be handled using new spectrum access algorithms in addition with spectrum hand-off capabilities of CR-WSNs.
C. Challenges of CR-WSN
In this section we are going to discuss about challenges that CR-WSNs face (Kaur, 2013):
1) Detection, False Positive, Miss-Detection Probability: Detection probability is a metric indicating the legitimate detection of white spaces by CR-WSN. Miss-detection is the metric showing how much of these white spaces are being missed by the CR-WSN. If we assign C as the assumption of the occupied portion of the network by primary users and W as the white spaces of the spectrum,
C: Currently Occupied Spectrum
W: White Spaces
2) Hardware: Unlike Traditional WSNs, CR sensor networks face some limitations regarding processing capabilities, memory and power consumption. Therefore, CR-WSNs should be able to alter restrictions and transceivers based on their surroundings. For example to resolve the power consumption issue, based on the environment, each sensor node can use one of the energy harvesting or battery recharging methods (Akan, 2009).
3) Topology changes: As mentioned previously, CR-WSN topologies, similar to conventional sensor networks can be Ad Hoc, Clustered, Heterogeneous and hierarchical, or Mobile. The challenge this area faces is that in contrast with traditional sensor networks, network topology in CR-WSNs is more likely to change dynamically due to network requirements. This dynamically changing topology strategy helps network to optimize the power consumption of the network as well as increasing the speed of its data communication. As Akan et al prove, the more hops and the less inter-hop distance routes have within the CR-WSN network, the more efficient network energy consumption becomes. Hence, a flexible self-configuring topology would function more efficiently rather than a static topology. It is worth mentioning that this area has not received much research attention.
4) Manufacturing Costs: There are numerous functional advantages CR-WSNs possess over traditional sensor networks, most of which require more sophisticated hardware and software implementation. More sophisticated hardware and software, requires more money. This raise in the costs is one of the most important disadvantages of the combination of CR with sensor networks despite all the advantages it causes to happen.
D. Security
CR-WSNs are more exposed to security attacks than traditional sensor networks due to lack of any partnership between primary and secondary users. This allows third parties to falsify information that CR is receiving as a feedback of its surrounding spectrums. Spectrum sensing data falsification (SSDF) is one of the most important threats that these networks have to face. This threat happens when the attacker tries to find the spectrum white spaces and send some data similar to the data regularly sent by PU, this attack can cause denial of service of CR if it happens in large scale. Further than the potential threats WSNs face (Akan, 2009), there are other security issues in CR-WSNs such as data modification, data injection, access to private data, etc.
VI. Conducted Researches and open research trends
There have been many researches undertaken around the usage of CR in WSNs. Although this research area is still in its infancy, it is experiencing a rapid growth. One of most significant researches in this field is the one conducted by Cavalcanti et al. They designed a theoretical CR oriented WSN operating in ISM band (2.4 GHz) and weighed that against the standard ZigBee WSN. The assumption of this research is that both networks have similar unit gain and receiver’s sensitivities for receiving antennas in both WSNs were set at -85 dBm. As a result of this comparison, authors found out if we use the equal transmission power for both WSNs, the new CR-based designed method doubles the maximum communication of the channel. The so-called doubled range results into resolving the hidden node problem partially and declines the number of hops each packet travels; hence the performance of Medium Access Control proves to be more efficient in the CR-based design.
Other researchers have conducted some research around this issue. For instance (Akan, 2009). Examine the application of CR in WSN alongside with the design principles and possible advantages and weak points of this design are as follow. They analyzed the opportunities and threats of using CR and WSN in these networks. Vijay et al proposed prospective, survey, and key technologies on CR-WSNs. Bicen (2012) considered the delay-sensitive and multimedia communication in CR-WSNs in a range of environments. Liang et al.
According to Akan, open research trends in CR-WSN field includes CRSN node development, Spectrum-aware clustering and node deployment strategies including dynamic spectrum-based optimal node deployment strategies in terms of network coverage, spectrum utilization, and power consumption in CRSN topologies. Also an optimum power amplifier is still an open area for research for energy-efficient cognitive radio sensor network.
VII. Conclusion and Further Work
A cognitive radio wireless sensor network is a type of sensor network that consists of spatially-scattered independent CR equipped wireless sensors monitoring the physical or environmental conditions concurrently. This research discussed the development of CR-WSNs, opportunities, challenges, structure, and research trends. Some of the recent research results in CR-WSNs were surveyed. CR-WSNs as a new technology have several areas remaining to be explored and enhanced. For the success of CR-WSNs, massive research is required in several aspects. Considerable developments in hardware, software and algorithms are required to make elegant cognitive sensors. The following are the potential challenges for the success of CR-WSNs:
- Detection, False Positive, Miss-Detection Probability
- Hardware challenges regarding processing capabilities, memory and power consumption.
- Topology change
- Manufacturing Costs
VIII. References
Akan, O., Karli, O., & Ergul, O. (2009). Cognitive radio sensor networks. Network, IEEE, 23(4), 34-40.
Ben Letaief, K., & Zhang, W. (2009). Cooperative communications for cognitive radio networks. Proceedings of the IEEE, 97(5), 878-893.
Bicen, A. O., Gungor, V. C., & Akan, O. B. (2012). Delay-sensitive and multimedia communication in cognitive radio sensor networks. Ad Hoc Networks, 10(5), 816-830.
Cabric, D., Mishra, S. M., & Brodersen, R. W. (2004, November). Implementation issues in spectrum sensing for cognitive radios. In Signals, systems and computers, 2004. Conference record of the thirty-eighth Asilomar conference on (Vol. 1, pp. 772-776). IEEE.
Cavalcanti, D., Das, S., Wang, J., & Challapali, K. (2008, August). Cognitive radio based wireless sensor networks. In Computer Communications and Networks, 2008. ICCCN’08. Proceedings of 17th International Conference on (pp. 1-6). IEEE.
Cayirci, E., & Rong, C. (2008). Security in wireless ad hoc and sensor networks. John Wiley & Sons.
Gardner, W. A., & Spooner, C. M. (1992). Signal interception: performance advantages of cyclic-feature detectors. Communications, IEEE Transactions on, 40(1), 149-159.
Ghasemi, A., & Sousa, E. S. (2005, November). Collaborative spectrum sensing for opportunistic access in fading environments. In New Frontiers in Dynamic Spectrum Access Networks, 2005. DySPAN 2005. 2005 First IEEE International Symposium on (pp. 131-136). IEEE.
Hoven, N., Tandra, R., & Sahai, A. (2005). Some fundamental limits on cognitive radio. Wireless Foundations EECS, Univ. of California, Berkeley.
Joshi, G. P., Nam, S. Y., & Kim, S. W. (2013). Cognitive radio wireless sensor networks: applications, challenges and research trends. Sensors,13(9), 11196-11228.
Kaur, S. (2013). Pushing frontiers with the first lady of emerging technologies-Intelligence in Wireless Networks with Cognitive Radio Networks!. IETE Technical Review, 30(1), 6.
Kim, S. W. (2013). Cognitive Radio Wireless Sensor Networks: Applications, Challenges and Research Trends. Sensors, 13(9), 11196-11228, pp. 1-8.
Uthra, R. A., & Raja, S. V. (2012). QoS routing in wireless sensor networks—a survey. ACM Computing Surveys (CSUR), 45(1), 9.IEEE Standard for Information Technology- Telecommunications and Information Exchange between Systems- Local and Metropolitan Area Networks- Specific Requirements Part 15.4: Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for Low-Rate Wireless Personal Area Networks (WPANs), IEEE Standard 802. 15.4-2006, 2006, (Revision of IEEE Std 802. 15. 4-2003).
Vijay, G., Ben Ali Bdira, E., & Ibnkahla, M. (2011). Cognition in wireless sensor networks: a perspective. Sensors Journal, IEEE, 11(3), 582-592, pp. 6-11.Joshi, G. P., Nam, S. Y., &
Zeng, Y., & Liang, Y. C. (2007, April). Covariance based signal detections for cognitive radio. In New Frontiers in Dynamic Spectrum Access Networks, 2007. DySPAN 2007. 2nd IEEE International Symposium on (pp. 202-207). IEEE.
