Syed Hashim Raza Bukhari1,2 Sajid Siraj1,3 and Mubashir Husain Rehmani1
1 COMSATS Institute of Information Technology, Wah Cantt, Pakistan
2 COMSATS Institute of Information Technology, Attock, Pakistan
3 University of Leeds, United Kingdom
The world is facing a huge urbanization and it is believed that around 70% of the world's population will be living in cities by 2050 Jin et al. (2014). As cities are growing, the need to provide smart solutions to every problem is increasing exponentially. The concept of ubiquitous applications has attracted people but on the other hand has raised a lot of problems for data communication technologies. In this section, we discuss the basics and motivation of WSN applications in smart cities.
With the emergence of smart cities, several technologies also need to be upgraded to become inline with information and communication technologies (ICT) required for smart applications Dohler et al. (2013). In this regard, the MAC protocol responsible for collision‐free communication for WSNs in smart cities is being revised Alvi et al. (2016). The concept of Internet of Things (IoT) for smart cities aims to exploit the most advanced communication technologies to support citizens in smart cities Zanella et al. (2014), Jin et al. ( 2014). For this purpose, general IoT‐based architectures are required. These architectures should be able to collect data from multiple applications being deployed in smart cities and then perform decision making based on that data Moreno et al. (2016). The decision making assists the system in performing tasks to facilitate the users. The architectures may be strengthened by the development of such softwares that can utilize dialogical logic between public administration and citizens. This facility can enhance the interaction between all the stakeholders for a certain task or decision Ortner et al. (2015).
As wireless networks are unreliable in nature, certain issues that may prevent the IoT to play its role are also required to be investigated Vlacheas et al. (2013). These networks also need to be energy efficient for seamless connectivity to peer nodes Zhu et al. (2013). The energy efficient networks encompass the concept of smart grids using ICT technologies Chen (2010). In this context, the implementation of IEEE802.11ah Wi‐Fi standard, which can provide connectivity between different devices, will play a vital role toward the maturity of smart cities Khorov et al. (2015).
The chapter has been organized in the following manner: Section 9.2 provides the role of WSNs in smart cities. In Section 9.3 we provide the basics and motivation of channel bonding based on CRSN‐based networks. Section 9.4 provides the future applications of channel bonding in CRSN‐based smart cities and the issues and challenges of their implementation have been highlighted in section 9.5. Finally the conclusion is in section 9.6.
WSNs have a key role in providing the information and communication technologies (ICT) in today's world. With WSNs, we can sense any event, transform it in the form of data and transmit it to any other place where it can actuate a motor, activate an alarm, etc. The applications of WSNs include home automation, traffic monitoring, health care applications, power system monitoring, and cab services, to name a few. In the context of smart cities, WSNs have become more important as now the deployed sensors are also involved in decision‐making activities. Now devices are talking to humans and making smart decisions just like humans. Figure 9.1 shows some applications of WSNs in smart cities. Traffic monitoring is one of the important factors of smart city management Celino & Kotoulas (2013). As the cities are growing, the traffic is growing as well. Now traffic can be handled from a central control room according to the traffic flow on roads Djahel et al. (2015). The system can now count all the cars present inside the city premises, number of cars entering or leaving the city etc. The audi connect project of audi automobiles will be available from 2017 models where a car can communicate with the central traffic control while entering any city and will get all information about situation of roads. The project will also enable the car to get the information about speed limits and traffic signals Audi connect technology 2017. Online: https://www.audiusa.com/technology/intelligence/audi‐connect (n.d.). It will help to avoid accidents in urban areas.
Underground transportation is a necessity of a smart city. It remarkably reduces the traffic on roads and provides connectivity in most parts of the city. However, the management of underground tunnels is also an issue. The underground tube management for efficient utilization of tracks as well as providing services to the passengers are required. These tunnels for underground transportation are also being utilized by other consumer services such as power lines, gas lines, sewerage lines, and backup links for data communication. The monitoring of all these services is essential for smooth operation of all services. Underground WSNs can remarkably support the operations of underground transportation and also can help the detection of any damage in a timely manner Akyildiz & Stuntebeck (2006).
Apart from underground transportation, smart cities should also provide smart cab services for citizens for traveling over short distances. In this regard, web‐based smart cab services are required, which can be located through a web‐based smart application. If a passenger needs a cab, he or she can use that application to locate any available nearby cab. Also, the company can get the status of all cabs in the city and can keep record of their traveled distance in a day. This facility will help the citizens to get the cab service quickly, and cabs can be utilized efficiently. The track record of passengers demand for a cab in certain area of a city will help the company to deploy specific number of cabs in that particular area Hu et al. (2015).
Table 9.1 Applications of WSNs in smart cities.
Applications | Ref | Year | Description |
Smart Traffic Management | Alvi et al. ( 2016) | 2016 | A TDMA based MAC protocol for WSN traffic to minimize the average packet delay in smart city environment |
Djahel et al. ( 2015) | 2015 | The role of WSNs for enhancing traffic management systems in smart cities | |
Shahidehpour et al. (2016) | 2015 | The future applications of WSNs in optimizing traffic signals in smart cities | |
Barone et al. (2014) | 2014 | WSNs based intelligent parking architecture for smart cities | |
Yoo (2013) | 2013 | The WSN based vehicle detection and evaluation system to enhance traffic condition in smart cities | |
Ruiz‐Garcia et al. (2010) | 2010 | A ZigBee based protocol to monitor the refrigerated food during transportation in smart cities | |
Boquete et al. (2010) | 2010 | A mobile system to analyze vehicle usage and billing for smart cities | |
Pérez et al. (2010) | 2010 | An RFID based fuzzy logic control to provide better control over vehicle for intelligent transportation in smart cities | |
Smart Grid Applications | Jiang et al. (2016) | 2016 | An algorithm for optimal placement of thermal sensor in smart grids |
Calvillo et al. (2016) | 2016 | Application of WSNs for energy management and planning in smart cities | |
Kylili & Fokaides (2015) | 2015 | The role of wireless sensors to achieve the target of zero energy buildings in smart cities | |
Magno et al. (2015) | 2015 | Smart LED lighting system to save energy and maintain user satisfaction in smart cities | |
Fang et al. (2012) | 2012 | A survey on new and improved smart grids for smart cities | |
Ma et al. (2011) | 2011 | Improving smart grids using cognitive radio for smart cities | |
Sreesha et al. (2011) | 2011 | A low latency routing support for smart grids using cognitive radio based WSNs for smart cities | |
Qiu et al. (2011) | 2011 | Implementation of cognitive radio network to support power and information flow in smart grids | |
Yi et al. (2011) | 2011 | Guidelines for ZigBee utilization for smart grid applications in smart cities | |
Gharavi & Hu (2011) | 2011 | A multi‐gateway structure to meet the requirements of smart grids for deployment in smart cities | |
Akyol et al. (2010) | 2010 | A survey of wireless communication techniques to support electric power system in smart grids | |
Hung et al. (2010) | 2010 | A linear network model to deploy WSN for communication in smart grids | |
Hochgraf et al. (2010) | 2010 | The utilization of GSM network for communication in smart grids | |
Farhangi (2010) | 2010 | The new paradigms and innovations in smart grids for next generation smart cities | |
Parikh et al. (2010) | 2010 | Challenges of wireless communication technologies for smart grid applications in smart cities | |
Gungor & Lambert (2006) | 2006 | The review of WSN deployments for electric system automation | |
Disaster and emergency applications | Gao et al. (2008) | 2008 | Implementation of WSNs for quick medical emergency response in smart cities |
Tseng et al. (2007) | 2007 | The exploitation of environment sensing capability of WSNs to improving citizens life in smart cities | |
George et al. (2010) | 2010 | The implementation of WSNs to develop a quick and accurate disaster response network for smart cities | |
Gray et al. (2011) | 2011 | A WSN based web architecture for integrating multiple heterogeneous data sets for environmental applications in smart cities | |
Casey et al. (2008) | 2008 | Deployment of WSN based tsunami detection and response system | |
Li et al. (2011) | 2011 | The implementation of WSN based water level monitoring system for smart cities | |
Smart health care applications | Villacorta et al. (2011) | 2011 | The implementation of WSNs for ambient assistance of senior citizens in smart cities |
Solanas et al. (2014) | 2014 | The application of WSNs to provide smart health solutions in smart cities | |
Chung & Liu (2013) | 2013 | A ZigBee based smart health care monitoring system | |
Yilmaz et al. (2010) | 2010 | A smart health care system using wearable physiological sensors | |
Morreale (2007) | 2007 | The application of WSNs to provide telehealth in smart city environment | |
Ko et al. (2009) | 2009 | A study of WSNs performance in emergency room of a smart hospital | |
Enabling Technologies | Khorov et al. ( 2015) | 2015 | The concept for applications of WSNs for providing enabling technologies in smart cities |
Mohammed et al. (2014) | 2014 | Applications of UAVs by integrating WSNs in smart cities | |
Miscellaneous Applications | Rashid & Rehmani (2016) | 2016 | A survey on applications of WSNs in smart cities |
Schleicher et al. (2015) | 2015 | The role of WSNs for next generation smart cities | |
Celino & Kotoulas ( 2013) | 2013 | The applications of WSNs for ubiquitous and pervasive solutions for smart cities | |
Dohler et al. ( 2013) | 2013 | The role of ICT based WSN applications in smart cities | |
Schleicher et al. (2016) | 2016 | The need of WSNs to address the requirements and architectural aspects of smart city application ecosystem |
The huge urbanization in cities proportionally generates waste due to daily life household as well as industrial activities. For this purpose, waste monitoring and cleaning requires an intelligent system to be able to sense the level of waste in any particular area of the city. The concept of green cities has emerged with the advent of smart waste monitoring and management systems. WSNs can be deployed to sense the waste level of waste bins in every apartment and can notify the company to pick the waste Longhi et al. (2012). The industrial waste should be taken more seriously as it is more concerned with the environmental health. The amount of toxic gases being added in the atmosphere should be monitored and limited to regulations imposed by the authorities. A variety of WSN applications in smart cities have been listed as Table 9.1.
The waste management companies and atmosphere health monitoring department can work in collaboration to maintain the clean environment of a smart city. The level of toxic gases in the atmosphere should be continuously monitored to keep the environment healthy. Moreover, the pollen count in the atmosphere, which can cause several allergies, should also be monitored Yu et al. (2013). The pollution due to fuel combustion in automobiles is a great concern of smart cities. The authorities have made regulations for the periodic checkups and maintenance of automobiles to minimize the pollution. Germany is the first country in this regard to ban all combustion engine vehicles by 2030 and converting all vehicles to electricity Online: http://www.roadandtrack.com/new‐cars/future‐cars/news/a31097/german‐government‐votes‐to‐ban‐internal‐combustion‐engines‐by‐2030/ (n.d.). It will reduce the demand of fuel but on the other hand, renewable energy sources are required to meet the high demand of electricity in the near future.
The high demand for electricity in smart cities can be handled by implementation of smart grids. Implementation of WSNs in smart grids make the supply and utilization of electricity efficient within the city. Now, the billing and complaints can be made using smart applications that simplify the processes and effectively generate the revenue for the supplier companies. The temperature of power lines can be monitored to estimate the power losses due to heat and any increase of temperature due to short circuits or any damage can be monitored Jiang et al. ( 2016), Fang et al. ( 2012).
Temperature and weather updates are frequently required for planning and development. For this purpose, weather forecasting departments work to timely inform the citizens about any emergency condition due to abrupt weather conditions. Web‐based smart weather forecasting applications keep the citizens updated about the coming weather conditions. By using this information, citizens plan their activities and prepare for any emergency condition Arampatzis et al. (2005).
The household operations of daily life activities have been made interactive with the help of WSNs Yang et al. (2009). Now the majority of these operations can be done with a single click as these applications have been associated with smart applications available in smart phones. Keeping the identity users as a key, the users can now operate these functions even when they are away from home and can monitor the activities of home appliances anywhere.
The structural health is an important parameter to monitor as it estimates the life of a building. It becomes more important when preservation of historical buildings is concerned. WSNs can be deployed to sense the vibrations making impact on the structures of historical places such as a museum. The data collected from these sensors can be used to estimate the structural health and to control the vibrations. The structural health of bridges can also be observed by this technology and maintenance work can be scheduled based on this data Cao & Liu (2016).
Next, we discuss the concept and importance of channel bonding to support large bandwidth applications.
Channel bonding is a technique to combine multiple contiguous channels to form a wide‐band channel Rehmani et al. (2012). The multimedia applications, which require high‐speed communication, can be implemented with the help of channel bonding. WSNs enriched with cognitive radio (CR) support can efficiently utilize the blank slots in licensed spectrum. This is due to the fact that ISM band is already overcrowded Steenkiste et al. (2009) and overlaid deployment of WSNs make the spectrum harder to access Lin & Chen (2014). The CR networks are the only viable solution to this problem Bukhari, Rehmani & Siraj (2016).
The motivation of channel bonding is to enhance the overall bandwidth by creating a bond of multiple contiguous available channels; hence the overall capacity of system increases with the number of bonding channels. Khan et al. (2014) has provided a detailed survey on the CB concept, which enables the opportunistic unlicensed users to coexist with the licensed users. It is required that channel bonding must avoid creating any interference with adjacent channels. For this purpose, guard bands should be introduced at the bond edges whereas the guard bands between the bonding channels will be assumed to be the part of the bond. It will provide almost 10% extra bandwidth as compared to channel aggregation. Moreover, CB does not incur additional overhead as control information is required only at the making and breaking of bond.
Channel bonding has been implemented in various networks such as traditional networks, which include cellular networks and wireless local area networks (WLANs), wireless sensor networks (WSNs), and cognitive radio networks (CRNs). In this section, we will discuss the role of channel bonding in improving the spectrum scarcity in smart cities.
The cellular networks with high‐speed communication services are adopting a channel‐ bonding scheme. Khan et al. ( 2014) have discussed the motivation to develop CB schemes to meet the high bandwidth requirements of next‐generation cellular networks. These CB schemes can be helpful for both licensed and unlicensed users to avail maximum benefits of resources. The recent shift of technology toward smart devices demands these devices to be energy efficient as well. As almost all of these devices are battery powered, increasing bandwidth requires compromising of transmission power to prolong the battery life. In this context, novel techniques are required that should provide high‐bandwidth data access while consuming low power. A data‐traffic‐aggregation‐based method for smart devices has been provided in Hu & Cao (2014), where an efficient search algorithm has been tested hence minimizing the energy consumption and delay in the network whereas a clustering‐based technique is given by Weber et al. (2014), which can be used to provide high bandwidth. Hence, CB schemes can be helpful in providing energy efficient solutions to the users. The recent shift of research toward high‐bandwidth applications has highlighted the importance of CB.
Channel‐bonding schemes have been considered as an effective solution to provide high bandwidth to meet the ever‐increasing demand of high‐bandwidth networks. The IEEE 802.11 specification for wireless local area networks (WLANs) proposes two techniques to combine the available channels for high bit rate. One is the multi‐channel technique, which enables a single node to simultaneously use multiple channels for communication, as seen in IEEE 802.11s, which defines how devices can connect making a wireless mesh network (WMN) and can be utilized by static or ad hoc networks 802.11s (2012). The multiple channels need not be contiguous in the multi‐channel technique. The other one is the CB technique, in which multiple contiguous channels can be combined together to make a single broadband channel, as seen in IEEE 802.11n (2009). Both techniques provide certain advantages along with limitations; for example, multi‐channel minimizes traffic congestion and on the other hand CB increases throughput of channel Xu et al. (2007). CB can utilize both licensed and unlicensed bands Khan et al. ( 2014) where contiguous channels can be found and assigned to CR nodes opportunistically. In wireless mesh networks, the use of multi‐channel multi‐interface Lim et al. (2011) can also be a good approach for increasing the throughput. In this scenario, neighboring nodes should not share the same channel in order to allow parallel packet transmission. When two neighboring nodes try to occupy the same channel, only one of them should transmit.
Conventional wireless sensor networks (WSNs) do not consider the opportunistic access; hence, all the nodes have the same priority over the network. In this situation, if the WSN traffic is low over the network, CB can be an effective approach to increase the bandwidth. Whereas, if the WSN traffic is high, CB may increase the number of collisions due to the few number of channels being shared between many nodes. All networks with fixed channel widths face a hard selection choice between transmitter range and power consumption. In case of WSNs, power consumption becomes a serious issue as the batteries have limited power installed at each node. Increasing the transmitter power may increase the communication range but at the expense of higher power consumption, eventually reducing the service life of the node. A possible answer to satisfy these constraints is to use CB that can dynamically adjust the channel width, when and as required. We can make a bond of contiguous channels when high throughput is required, and the bond can be broken after data transmission is completed. Chandra et al. (2008) has shown the method of adaptive channel width allocation in IEEE802.11b/g, which can be adapted for WSNs keeping in view the constraints specific to the WSNs. Another dynamic CB scheme has been proposed by Huang et al. (2013) that focuses on multicarrier wireless networks. The proposed scheme can perform well in WSNs and play an important role in reducing contention in wide‐band spectrum sharing and achieving fast spectrum agreement. However, to quantify the performance of any protocol, specific metrics need to be chosen depending upon the type and requirement of that particular protocol.
As far as CB is concerned, throughput is indeed the most widely used performance metric by researchers. It reveals the purpose of CB to enhance achieved data rate in WSNs. Chandra et al. ( 2008) has done a comparison on the impact of CB and proves that throughput over a channel increases with channel width but number of users and distance between the nodes play an important role to estimate throughput. The combination of throughput, number of nodes, and the distance between the nodes has collectively been called load balancing. Other works such as Sankaranarayanan et al. (2005), Kyasanur et al. (2005), Kim & Ko (2007), Cheng et al. (2006), Chiu et al. (2009), Sharma & Belding (2009), Pelechrinis et al. (2010), Gummadi et al. (2008), Deek et al. (2011) also have focused on throughput while studying CB.
The underutilized spectrum by cellular networks can be effectively used by wireless sensor nodes Sankaranarayanan et al. ( 2005) using dynamic spectrum access (DSA) approach and CB scheme enhance throughput of nodes. Moscibroda et al. (2008) has discussed the problem with fixed channel assignment that it does not effectively utilize the spectrum. In contrast, using dynamic spectrum allocation can assign the vacant slot to sensor nodes which can effectively utilize the channel by using CB. In the same fashion, Zhou et al. (2008) has proposed a spectrum auction scheme (VERITAS), which assures that the spectrum is efficiently utilized and assigned to those users only who have ability to access the channels.
Delay is a fundamental parameter while considering the performance of any network. Embedding a CB scheme in a network may have a significant impact over propagation delay. A user can experience more delay if there is large number of contending nodes in the network Sankaranarayanan et al. ( 2005), however in case of less traffic, the propagation delay can be minimized by using a CB scheme. Chiu et al. ( 2009) has revealed in his work that non‐overlapping channels can be used simultaneously, which will not only enhance throughput but will also reduce end‐to‐end delay.
CRNs are generally composed of two types of nodes: primary radio (PR) nodes and the CR nodes. Those users who have valid license to use the band are usually termed as primary users Popescu (2012). They have a priority to access the channel over those users that do not have license –, generally called the CR users or secondary users. CR nodes can operate in licensed bands as well as in the unlicensed bands.
In the literature, various models have been used to imitate the activity of PR nodes, such as a Bernoulli process Banaei & Georghiades (2009), Markov renewal process (MRP) Pyke (1961), and deterministic process Chronopoulos et al. (2008), M/M/1 Malanchini et al. (2009), M/G/1 Shiang & Schaar (2008). Out of these models, the MRP model has been widely used (See Lee & Akyildiz (2008), Yuan et al. (2007) for details). The ON/OFF PR activity model approximates the spectrum usage for CRNs. MRP has also been used for IEEE802.11b/g (WiFi) Geirhofer et al. (2006), voice networks Adas (1997), Sriram & Whitt (1986) and the patterns of communication in public safety band Zhe & Sana (2011), Rehmani (2011). Consider a frequency channel that remains occupied for 3 minutes on average and then remains unoccupied for the next 2 minutes on average. In such a scenario, when a CR node wants to communicate, the probability of finding the channel vacant is highly dependent on the current state (i.e,. whether vacant or occupied) and the total duration of renewal (i.e., going back to the vacant state). MRP assumes an exponential distribution for both the vacant and occupied states of the channel. MRP is applied where most applications involve processes in which a transition from a state to itself is possible, and these states should be finite Pyke ( 1961).
The PR activity patterns are generally categorized into four major types: long‐term, high, low, and intermittent activities. The long‐term activity is generated by those nodes that have long ON and long OFF periods. This type of activity can be seen in the cellular networks scenarios where users are subscribed to special packages, e.g., free call packages. In high‐PR activity, the channel has long active time but much shorter OFF periods. This can be seen in a highly congested urban area where all the channels are mostly occupied. In low PR activity, the channel has short ON and long OFF periods. This type of PR activity can be observed in remote areas or during less peak hours. In the intermittent activity, the channel has short ON and short OFF periods. This type of PR activity can be observed where users use the channels for very short periods of time, e.g., at bus stations or railway stations Rehmani et al. (2011).
These days most of the wireless equipment are using the industrial, scientific, and medical (ISM) radio band for communication. Due to the humongous increase in the Wi‐Fi‐ and Bluetooth‐enabled devices, the ISM band of 2.4 GHz has started becoming overly saturated. This is the reason that, nowadays, unlicensed traffic has been allowed to operate in the TV broadcast range. The spectrum dedicated for TV broadcasts can be used for data communication in a cognitive way Yuan et al. ( 2007). However, it has introduced certain new challenges: first, the CR nodes should be aware of TV broadcast traffic (through spectrum sensing); second, the CR nodes should be able to use the available bandwidth in an intelligent and efficient manner.
When dealing with multiple channels in CR network, channel assembling technique can get better performance as compared to unassembled channels Jiao et al. (2012). The effectiveness of dynamic spectrum aggregation has been studied in Jiao et al. (2011) and results prove that dynamic schemes can achieve higher throughput than randomly combining the idle frequency slots and assigning them to contending CR nodes. One main concern, while enjoying the higher throughput, is the increased complexity. The dynamic schemes are more complex as the required number of handshakes is increased in dynamic CR schemes Indika A. M. Balapuwaduge & Li (2012).
The concept of PR and CR nodes in a network make CB difficult to implement. Channel sensing and selection schemes (as discussed earlier) play vital roles to provide opportunistic access to CR nodes. To access the spectrum while minimizing the interference caused by secondary users, two approaches have been discussed in Pollin (2007), i.e., spectrum underlay and spectrum overlay. In spectrum underlay, the CR nodes can communicate within a certain threshold to avoid interference, but it limits the role of CB as maximum benefits of wide band cannot be achieved. Spectrum overlay limits the transmit power of secondary users so as to minimize the interference. However, the worst case is assumed for this technique that primary users will be communicating constantly. Spectrum overlay uses the concept of spectrum pooling in which cognitive radios seek spectrum holes for communication. Within such a hole, no restriction of transmission on secondary user is imposed. As discussed earlier, performance evaluation is necessary to check the effectiveness of CB. As in CRNs, the concept of PR and CR node activity exists, so one cannot ignore these parameters while analyzing CRNs. The higher the number of active PR nodes at any given time, the lower the probability of contention. Therefore, the networks having low PR activity are considered to be suitable for CR nodes as higher throughput is achievable in such networks. If we assume that there is no PR activity, only the CR nodes will be contending for a channel, and will determine the throughput and bond size.
In CRNs, the CR nodes have to first sense the channel and in case of presence of PR activity, they identify the type of PR activity present on the channel. The type of PR activity is important for CR nodes to apply CB. Let us assume that when a CR node senses a particular channel, low PR activity is identified. This type of PR activity is suitable to apply CB as there will be short ON and long OFF periods on the channels. These long OFF periods can be utilized by the CR node for CB. In a scenario where the CR node detects short‐term PR activity, it will be having short ON and short OFF periods. This type of PR activity is not suitable for CB as there will be very short periods available for CR nodes and very frequent bond establishment and termination will be required.
For CB in CR‐based networks, throughput has a trade‐off with the number of contending CR nodes at any given time Su & Zhang (2007). This is the reason why most of researchers have selected this parameter for performance optimization. Cordeiro, Challapali, Birru & Shankar (2006) discusses IEEE802.22, the standard based on cognitive radios support DSA that can futuristically be used for CB. Auer et al. (2007) has reviewed a DSA‐based approach as throughput‐delay relationship in which maximizing one reduces the other. For efficiently utilizing the benefits of CB in cognitive networks, a throughput efficient scheme is presented by Lu et al. (2009), which divides the spectrum in slots so that traffic over these slots can be sensed correctly and when slots are found idle, they can be bonded for increasing bandwidth. Throughput has also been used as key factor for indicating performance by other researchers in the literature Jia et al. (2008), Bian & Park (2007), Talat & Wang (2008), Salameh et al. (2013), Cordeiro, Challapali & Ghoshr (2006), Geirhofer et al. (2009), Rahul et al. (2008), Yau et al. (2009), Jiao et al. ( 2011).
As discussed earlier, PR user activity has a large impact over the spectrum utilization by CR nodes as well as other CR‐related performance parameters also rely on the activity pattern of PR nodes Saleem & Rehmani (2014). The CR nodes has to wait unless PR nodes vacate the spectrum Talat & Wang ( 2008). This time duration affects the performance of CR nodes, which are continuously sensing the spectrum for utilization. Hence, a low PR activity is desirable for CR nodes to maximize the opportunistic spectrum utilization Joshi et al. (2012). The concept of cognitive femtocells has been proposed by Gur et al. (2010) so that spectrum efficiency can be achieved in a smaller region, and chances of opportunistic access increases as PR traffic is spread over a large region as compared to smaller cognitive femtocells.
CB in CRSNs has to cope with two issues Azarfar et al. (2012); firstly, it has to use low transmit power due to the lifetime of the power‐constrained sensor nodes, and secondly, it has to provide maximum capacity gain to CR nodes. Hence, we can say that we have to take all the issues of CRNs and WSNs while implementing an efficient CRSN. The efficient radio resource allocation in CRSNs is important due to the dynamic channel selection of CR nodes. A comprehensive survey on radio resource allocation in CRSNs has been presented recently in Ahmad et al. (2015). The survey covers the existing resource allocation schemes along with probing unexplored research directions such as inter‐network interference and cross‐layer resource optimization.
In the need for larger bandwidth, CB in traditional wireless networks is performed at the data‐link layer Cordeiro, Challapali & Ghoshr ( 2006). The data‐link layer protocols, like media access control (MAC), usually establish a central node for decisions regarding channel selection and bonding. Although these schemes work quite well in traditional networks, the concept of self‐organizing and self‐healing nodes makes them inappropriate for CRSNs Akyildiz et al. (2002). These protocols do not address the issues raised by topological changes and scalability, which can be addressed if channel selection and bonding are done at the network layer.
Bukhari et al. (2016) have proposed a novel scheme for channel bonding in cognitive radio sensor networks Bukhari, Siraj & Rehmani (2016) and shown that channel bonding can be helpful in providing a large bandwidth to opportunistic users.
Smart Cities contain a vast range of applications, which contain multimedia content and require high bandwidth. Since, CRSNs operate in unlicensed as well as licensed bands, channel bonding can be a good approach for providing wide band for high‐speed communication. These applications can be smart traffic monitoring and management as cameras with speed guns take snapshots of vehicles exceeding the speed limit and send it to a central control room in real time so that the vehicle should be charged before it leaves the city premises.
Another scenario can be smart health care applications. These applications are very famous these days specially for senior citizens. Senior citizens with serious health issues use real‐time smart health monitoring systems and in case of any emergency, the attributes of patients' conditions, which include health parameters and patients' locations, should be communicated to their physicians. The physician can respond only if the information about the patient is received in time. The health care monitoring system can also invoke the ambulance service if the patient needs to be taken to the hospital. All this activity requires high‐speed communication, and channel bonding can be very helpful for this purpose Solanas et al. ( 2014).
An important feature of smart cities is machine‐to‐machine (M2M) communications, which makes the processes faster and more reliable. M2M communications techniques make use of long‐term evolution advanced (LTE‐A) cellular networks Dohler et al. ( 2013), and CB can be a more dynamic approach to provide high‐bandwidth communication. M2M communication takes input from various CRSNs and utilizes this information for decision making and activating the response system.
As smart cities contain a huge number of WSN‐based smart applications, the multiple WSNs deployed in the same area can cause considerable interference with each other. The CRSNs contain the solution, and now multiple concurrent deployments of CRSNs are possible as the CRSN node follows cognitive cycle to access the free channel. Moreover, the CB scheme can enhance the throughput for high‐bandwidth applications. To summarize, now citizens can benefit from using multiple WSN‐based applications with high‐bandwidth requirement within a room or a building.
A smart home is a building block of smart cities. The sensor nodes can be embedded into home appliances. Smart homes containing interactive applications such as smart metering, smart security features, smart lighting, smart kitchen, and smart washing applications highly require a CRSN‐based CB method to communicate with the users simultaneously along with providing the required performance. The users can also manage and monitor these home appliances easily both locally and remotely.
Smart environmental control provides the application of WSNs to keep track of pollutants in the atmosphere, smart waste management, birds movement detection, animal activities detection, weather conditions update, climate control, flood detection, etc. All these applications can benefit from CRSNs to communicate their data to the main server. As these sensor can be very large in number and access points can be installed to relay the data in a multi‐hop manner, access points can utilize CB to relay the collected data to the servers.
The IoT‐based cognitive solutions are the fundamentals of smart cities operations Vlacheas et al. ( 2013). Our proposed solution of channel bonding based on CRSNs will be helpful in providing the solutions for technological barriers where CRSN nodes can send data (through CB) to their licensed nodes, which can then utilize high speed Internet for efficient data delivery in smart cities.
The applications being deployed in smart cities are bandwidth hungry due to user‐friendly multimedia content. As discussed earlier, CB is an approach to provide larger bandwidth to users, and it can be easily adopted in smart city scenarios as CRSN nodes can utilize any available frequency band.
In this section, various issues and challenges regarding implementation of channel bonding in smart cities have been discussed. These issues will be helpful in opening future research directions in the field of channel bonding in smart cities.
Smart cities must ensure the basic rights of citizens. Citizens encounter multiple units that collect their information such as smart phones, public computers, and data collection centres. The huge amounts of data collected are stored in the cloud and become a hot target of hackers. So the issues of privacy and data security need to be addressed. Martinez‐Balleste et al. (2013) has listed these issues as a 5‐D model. The issues can be, but are not limited to, identity privacy, query privacy, location privacy, footprint privacy, and owner privacy. It is the responsibility of regulatory authorities to ensure that the information of citizens is safe from hackers and intruders. As CB is the solution to high‐bandwidth applications and smart devices may need to perform CB frequently, the privacy of citizens must be taken as primary objective especially while accessing the unlicensed band.
Smart cities offer a variety of ICT‐based solutions to the citizens. However, it is required that these solutions must also be energy efficient. The energy efficiency can be obtained by several methods such as complexity, density, and power‐saving policies CB‐based applications are focused on providing high‐speed communication by changing the channel width. As CRSNs contain wireless sensor nodes, these solutions must be energy efficient for providing seamless high‐speed services to citizens for longer periods of time Zhu et al. ( 2013).
In smart cities, there are numerous data sources at many locations. A variety of sensors are distributed throughout the cities. These sensors and data sources can be in the form of wearable sensors, smart cards, or vehicles. As all these sensors collect huge amount of data from a number of IoT‐based services provided to citizens, it is expected that storage and aggregation of data must be critically ensured. In this regard, big data‐based techniques are required, which should be able to handle and analyze the data along with providing quick responses to queries Moreno et al. ( 2016).
Cloud‐based services provide data centers for big data and geographically distributed locations. However, it is required that cloud‐based applications should be aware of which components have been deployed in which data center. It will help to minimize the cost and maximize the efficiency of CRSN‐based smart services Schleicher et al. ( 2016). The cloud‐based services should also be flexible as clouds are highly dynamic and the number of user activities and their demands may vary rapidly.
Interference occur when two wireless nodes try to access the same frequency channel at the same time. As smart cities contain multiple radio technologies deployed concurrently in the same geographic location such as IEEE 802.11 (Wi‐Fi) and IEEE 802.15.4 (ZigBee), it is further envisioned that cognitive radio technology efficiently senses the spectrum and utilizes it accordingly. It is required that more sophisticated and less complex techniques should be developed to avoid interference and enhance the spectrum utilization Avelar et al. (2015).
In this chapter, we have provided the motivation to implement the WSN‐based solutions in smart cities. Moreover, we have envisioned the futuristic cognitive radio sensor networks‐based channel bonding technique to meet the high bandwidth requirement of smart city applications. We have discussed that CB technique can be a good approach for high bandwidth applications in smart cities. We have also highlighted some directions for future research for better infrastructure deployment and utilization.
18.219.95.244