Minimizing Interference in Wireless Mesh Networks Based Telemedicine System

Problem statement: It has been always a challenging task for the resea ch rs to incorporate information technology advancements in the medical profession. In order to guarantee a reliable network infrastructure that supports the employment of wireless technology and mobile devices, a hybrid infrastructure that integrates wireless netw orks with the traditional wired Local Area Networks (LANs) is needed. The wired network will provide th e stability to the network, improve the overall performance of the hybrid network and provide an al ternative to the mobile routes, hence increasing the degree of redundancy and reliability. In recent years, Wireless Mesh Networks (WMN) have become more popular than ever. WMN has advantage of robust fault tolerance. Even if some of the mesh nodes are incapable, there exist many other al tern tive nodes to serve relay. In addition, multihop of WMN can not only spread coverage but also sa ve both cabling cost and human resource. Health care can drastically benefit by incorporating techn ological advancements, particularly wireless mesh technology. In this study we have presented a telem dicine system for rural and urban scenario based on WMN so that the medical community gets benefited by state of art broadband connectivity. Also in this study we have addressed a critical issue of pe rformance degradation in WMN due to interference and varying load. Approach: For this we have proposed a routing protocol AODV L BIARM. We have incorporated our routing metric Load Balancing I terference Aware Routing Metric (LBIARM) in AODV protocol to minimize the interference and c ater for varying traffic load in multi hop WMN based telemedicine system, thereby maximizing netwo rk throughput. Results and Conclusion: This protocol is evaluated by simulating in Opnet Modele r 16.1 PL1. We have found that our proposed protocol performs better than AODV WCETT and AODV H OP-COUNT.


INTRODUCTION
Telemedicine system: The word telemedicine is made up of two words. The first word, "tele," means "at a distance," and the second word, "medicine," is derived from the Latin ars medicina, meaning the art of healing. So medicine can be defined as the science and art of healing. It encompasses a variety of health care practices evolved to maintain and restore health by the prevention and treatment of illness. Combining the both telemedicine can be defined as the practice of medicine at a distance. One of the current accepted definitions of telemedicine is as follows.
Telemedicine involves the use of modern information technology, involving two-way interactive audio/video telecommunications, computers and telemetry, to deliver health services to remote patients. It facilitatec information exchange between primary care physicians and specialists at remote distance from each other.
There are two modes of operation for telemedicine: real-time and store-and-forward. In real-time mode, the information, which can be any combination of audio, video, data and images is transmitted to a remote terminal immediately after acquisition, thus allowing real-time interaction between patients and health care personnel. Some examples are consultation with a remote doctor, ambulatory Electrocardiogram (ECG) monitoring and instant transmission of ultrasound images. Because real-time applications involve continuous, time-critical information exchange, high transmission bandwidths are often required, resulting in relatively high operation cost. In store-and-forward mode, the information gathered is analyzed at a remote  (Akyildiz et al., 2005) is a promising technology, which provides wireless broadband connectivity to the Internet. WMNs are self-forming, self-healing and self-organizing. Their easy configuration and deployment make them an economical, reliable and simple solution that can be implemented anywhere at any time. The end users can use WMN to access the internet by connecting to any of the routers. WMNs have emerged as a fundamental technology for the next generation of Wireless Networking. They are formed by mesh routers and mesh clients as shown in Fig. 2. WMN are being investigated at this moment to support video-related applications such as video streaming, multimedia messaging, teleconference, voice over IP and video telemedicine due to its superior characteristics compared with other wireless standard, including mobility, high data rate and low cost infrastructure. With such infrastructure in place, a number of critical medical problems can be addressed with the help of WMNs and the stability of the infrastructure backbone can be done.
In a wireless mesh network, the capacity is reduced by interference from simultaneous wireless transmissions. Two types of interference affect the throughput of multi-hop wireless networks: intra-flow interference, which is induced between adjacent nodes on the same routing path and interflow interference, which is caused by nodes from neighboring paths. To provide backbone support to the IEEE 802.11 mesh based telemedicine system, it is necessary to reduce interference in WMNs which is a critical issue.
Interference: Interference is an important factor which influences the ability of two linked nodes to reliably communicate with each other at the desired communication rate. Within the scope of 802.11 based wireless networks, we distinguish between two types of interference: radio interference and channel contention interference. Radio interference represents a physical interference that influences the entire spectrum of electromagnetic waves. Physical interference has two distinctive subtypes: destructive interference and constructive interference. Destructive interference is depicted in Fig. 3a where 3 waveforms can be seen: the first waveform on the top (in red) represents message signal, the second waveform in the middle represents an unwanted interference signal and the bottom waveform represent the resulting signal waveform. As it can be observed in Fig. 3a, the destructive interference has a detrimental effect on the transmitted radio signal, which often results in a loss of transmitted data. On the other hand, in the second type of interference i.e., the constructive interference depicted in Fig. 3b (waveforms are in the same order as in Fig. 3a has the converse effect on the signal as it results in an increase of its amplitude. However, the waveforms used in these two examples are synchronized in time, which is very rarely the case in reality. This means that in most of the cases both forms of interference have a detrimental effect on the signal. The second type of interference i.e., the contention interference is not a physical interference as in the previous cases although it also produces a negative effect. Contention interference stems from the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) based MAC layer of 802.11 protocol that defines the behavior of the 802.11 station, which has to wait until a channel is free to commence its transmission. Therefore, the channel may be occupied by transmission from any node that uses the same channel within the communication range. Due to the license free availability of 802.11 band, this issue is even more pronounced in urban areas that contain a large number of 802.11 based networks. Thus, in urban areas all the co-located networks compete for the use of limited number of channels. In addition to the term interference it is important to clearly define two related terms: communication range and interference range. Communication range is the range in which a reliable communication between two nodes is possible and interference range is the range in which transmissions from one node can detrimentally affect the transmissions from other nodes on a same or partially overlapping channel. It is important to note that the interference range is always larger than the communication range as shown in Fig. 4. Contention Interference can further be classified.
Intra flow interference: Intra flow interference occurs due to the adjacent nodes on the same routing path. They compete against each other for channel bandwidth. This intra flow interference causes throughput to degrade severely due to the consumption of the flow bandwidth across each node on the same routing path. The hop count of the flow increases with an increase in end to end delay. This causes congestion. E.g., in Fig. 5, we see the path 1→2→3 causes intra flow interference because of the reuse of channel 1 on the flow 1→2 and from 2→3. So the path 1→4 →3 does not have intra flow interference due to the assigning of two different channels between 1→4 and 4→3. We can say that that 1→4→3 is a better path in comparison to 1→2→3. So a good intra flow aware interference matric should assign 1→4→3 a lower weight than 1→2→3. In other words we can say that a good routing metric reduces inter flow interference by selecting nonoverlapping channels for adjacent hops of a path.

Inter flow interference:
Inter flow interference is the interference occurring due to the other flows operating on the same channels and competing for the medium. This is caused by the multiple flows between different routing paths as shown in Fig. 6. This not only consumes bandwidth of the nodes along its route, it also competes for bandwidth with the nodes following in the neighborhood. In comparison to intra flow interference inter flow interference is harder to control due to the involvement of multiple flows and routes. In Fig. 6 we see inter flow interference due to the two paths namely 1→4 →3 and 5→6→7. A good inter flow aware metric should assign a low weight1→2→3 then to 1→4 →3 as path 1→2→3 has less inter flow interference.  Malan et al. (2004) had proposed an ad hoc sensor network infrastructure for emergency medical care. Chu and Ganz (2004) had presented a mobile teletrauma system based on 3G networks. Golmie et al. (2005) presented the performance evaluation of WPANs for sensors and medical application. Hop count is the most common routing metric used in Dynamic Source Routing (DSR) (Johnson and Maltz, 1996) and Ad-hoc On-demand Distance Vector (AODV) (Perkins et al., 2003) routing protocols. Hop count metric is not suitable for WMN due to parameters like congestion control, scalability and in establishment of paths with maximum throughput. The Expected Transmission Count (ETX) routing metric was proposed by Couto et al. (2003) to address the above mentioned problem. ETX does not take account of load and intra flow interference. It is insensitive to various link rates. The Expected Transmission Time ETT (Draves et al., 2004) routing metric was an improvement over ETX as it took into account the bandwidth of different links. The disadvantage of this metric was it did not consider link load explicitly and as a result was not able to avoid routing traffic through already congested nodes and links. The second disadvantage was it did not minimize intra flow interference as it was not designed for multi radio networks. The Weighted Cumulative Expected Transmission Time (WCETT) (Draves et al., 2004) was designed to improve the ETT metric in the area of multi radio mesh networks by considering channel diversity. A number of routing metrics have been proposed for routing protocols in Wireless Mesh Networks such as metric of interference and channel switching (MIC) (Yaling et al., 2005), Load Aware ETT (LAETT) (Aiache et al., 2008), Exclusive ETT (EETT) (Jiang et al., 2007), Interference Load Aware(ILA) (Shila and Anjali, 2007), interference Aware iAWARE (Subramanian et al., 2006), Adaptive Load Aware Routing Metric (ALARM) (Amir et al., 2009) and a location Aware Routing Metric (ALARM) (Alotaibi and Roy, 2008) were proposed to support load balanced routing and to consider intra and inter-flow interference, in addition to being isotonic. We had found that each routing metric had limitation in not meeting one or more criteria. We had proposed a new routing metric called Load Balancing Interference Aware Routing Metric (LBIARM) (Siraj and Bakar, 2011b) which takes cares of all the above parameters.
AODV LBIARM routing protocol: We incorporate our LBIARM metric in AODV protocol. This protocol finds best route discovery. It selects the route which has minimum interference, high throughput and low end to end delay. The route discovery is based on the following algorithm.

Algorithm for best route discovery:
Source Node S Broadcasts to Gateway Node G if (Intermediate Node I receives Route_Request) then Calculate path load from S to I using the following equation: Performance Evaluation: Simulation environment: The performance evaluation of our proposed metric LBIARM was done with hop count and WCETT metric by incorporating these metric in AODV protocol and simulating in Opnet Modeler 16.1 PL1. The network topology as shown in Fig. 7 was used for simulation. It consists of 16 static mesh routers uniformly randomly placed in a 4×4 grid in a 800×800 m 2 area. The average distance between each pair of two one hop nodes is the same. The interference range is set to be approximately equal as all mesh routers are with similar transmission powers. The source nodes send Constant Bit Rate (CBR) traffic with UDP as transport protocol, consisting of 1024 byte packets with a sending rate of 20 packets/second. The center mesh router was selected as the Gateway node. The value of was taken as 0.3 for Calculation of WCETT and LBIARM. This value was taken as it was found that optimum value of WCETT is when a = 0.3. The following simulation parameters were used (Table 1).
Performance metrics: For evaluation of AODV LBIARM the following metrics were used.
Packet delivery ratio: This is the ratio between the number of data packets successfully received by the destination node and the total number of data packets sent by the source node. This metric reflects the degree of reliability of the routing protocol. End-to-end delay of data packets: This is the delay between the time at which the data packet originated at the source and the time it reaches the destination and includes all possible delays caused by queuing for transmission at the node, buffering the packet for a detour, retransmission delay at the MAC layer, propagation delay and transmission delay. This metric reflects the quality of routing protocol.
Throughput: This is defined as the amount of data that is transmitted through the network per unit time (i.e., data bytes delivered to their destinations per second).
Simulation results and analysis: Packet delivery ratio: Figure 8 shows the percentage of packet delivery ratio between LBIARM and WCETT and HOP-COUNT in presence of interfering traffic. We observe with the increase in flows contention increases. With an increase of interference, the packet loss ratio of all the above metrics also increased. Percentage of Packet delivery also falls down. It is observed AODV LBIARM performance is better than AODV WCETT and AODV.
End to end delay: Figure 9 shows the end to end delay .It is observed AODV LBIARM performs better than AODV WCETT and AODV. This is due to the balancing act of LBIARM which minimizes interference due to the selection of an optimum route. This route is based on less interference taking into consideration of varying traffic load. When the flows are increased it increases the end to end delay. LBIARM improved the end to end delay as AODV HOP-COUNT and AODV WCETT do not balance the traffic flows and create congestion causing large delay. Throughput: Figure 10 shows throughput comparison between AODV, AODV WCETT and AODV LBIARM. The throughput in AODV LBIARM is greater because the other two metrics creates congestion areas. At each hop there is a delay of data packets. AODV LBIARM creates quality links with fewer delays which are less loaded.

CONCLUSION
In this study we have proposed a telemedicine system which utilizes Wireless Mesh Networks as a backbone to utilize the benefits of Broadband networking in telemedicine. This proposed scenario can be applied to rural as well as for urban areas. A common problem in WMN is the performance degradation with multiple hops. This is due to Interference of the neighbors and varying traffic load. This is a critical issue in WMN. We had incorporated our proposed LBIARM metric in AODV protocol to address this problem. Our metric has the ability to cater for varying load and capture inter and intra flow interference. Also it had the ability to capture difference in transmission rates. We have shown by simulations that AODV protocol incorporated with LBIARM metric performs better as compared to WCETT and HOP-COUNT routing metric. It will be interesting to know the performance measurements if implemented in a test bed. Then the practical feasibility of LBIARM would be known.