802.11p Profile Adaptive MAC Protocol for Non-Safety Messages on Vehicular Ad Hoc Networks

,


INTRODUCTION
Vehicular Ad Hoc Networks (VANET) is a continuosly imperative field in Mobile Ad Hoc Networks (MANETs). Lately, researchers have demonstrated an increased interest in Vehicular communications. As indicated by authors in [1], the VANETs consist of vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communications derived from wireless LAN technologies. The typical set of VANET application (e.g., cars accident triggering and local restaurant information for drivers), possessions (driving licensed, rechargeable power outlet), and the surrounding (e.g., highway traffic coditions, security concerns) make the VANET a distinctive area of wireless communication.
The demanding part of VANET is the high mobility of vehicles which influences the high rate of topology changes and the high inconsistency of node intensity. It can be presumed that MANET routing protocols are hard to instigate, eg. Obsolete neighbor data in routing table protocol [2]. The focal point of this study was on value-added applications that were classified as on-request services identified as infotainment, interative media or non-safety applications. Notification triggers for tourists such as hotel and lodging, erestaurants or e-map downloading shorten the time and thus save fuel utilization.  Figure 1 represented the highway driving environments which clarified the three lanes for each driving direction. These traffic congestion and road safety information can be distributed through various service center in different RSU clusters for broadcasting. Each vehicle contributed in detecting and updating most recent street data. Web correspondence on vehicles, information exchange between OBU (on board unit) in vehicles and RSU (road side unit), depend on wireless remote systems. There are three wireless radio technology measurements set for the vehicle communication [3] [4]. The FCC partitioned the spectrum into seven channels, each with 10 to 20 MHz, in which six were recognized as Service Channels (SCHs), and one as a Control Channel (CCH). The CCH channel is utilized for safety messages while non-safety services (WAVE-mode short messages) adhered to six other SCH service channels that were accessible [5][6]. Table 1 summarized and identified the distinctive functionality of the 802.11 standard in WLAN operation. The 802.11b and 802.11g protocols have been used substansively, which then were revised followed by 802.11n and 802.11p. The 802.11p is another multi-streaming modulation technique. The WLAN standard worked on the 2.4 GHz and 5 GHz Industrial, Science and Medical (ISM) frequency bands.  Table 2, numerous standards could be accomplished associated with wireless local area network (WLAN) in VANETs availability. There are different functionalities among all the standards. Security specification, routing, addressing services and interoperability are protocols that affect RSU equipments and OBU communication.

CONGESTION CONTROL ALGORITHM DESIGN CRITERIA
The congestion control is complicated due to increase in traffic and data exchange. Plenty of application exchange the data from one user to multiple users, resultant the congestion control became complicated [36]. Congestion control algorithm can be embraced from numerous methods to reduce congestion in VANETs. To avoid congestion, some are depend on broadcasted messages to its neighbors. Table 3 demonstrates the most contemplated congestion control algorithms and parameters by many researchers based on broadcasting cautioning messages.  [22] Throughput improvement at SCHs and transmission delay reduction Dynamic carrier sense threshold [23] Transmission power and packet generation Multichannel communications [24] Multichannel architecture enhancement Dynamic service-channels allocation (DSCA) [6] Multiple service-channels based on a single transceiver VEMMAC protocol in VANETs [25] Multi-channel MAC enhancement

SIMULATION AND TESTING PHASE
Experimenting and testing vehicular network requires intensive labor and high expenses. Hence, an alternative solution is to use the simulation before actual implementation [34]. In this project, test-bed efforts were done utilizing OMNeT++ ver 4.6 simulator [28] running under UBUNTU 14.04.2 LTS.
All medium access control (MAC) and routing protocols depended on the INET framework [29], [30] and INET-MANET [31] of the OMNeT++. The mixture of control factors and noise factors are demonstrated in Table III as the experiment parameters. The simulation time for each experiment parameters was 250 seconds and 3 (RNG) random seed generation were conducted [27]. This research optimized the control factors in VANET congestion controls to attain least end-to-end delay, maximum PDR and maximum throughput for highway test condition environment. Simulation parameters for the tested experiments are as expressed in Table 4.
The packet sizes utilized were 25KB up to 125KB and there were five location of RSU distance along the highway. 802.11p was utilized for MAC protocols while the routing protocol AODV was chosen. The control factors level of variations are expressed in Table 5. The noise factors levels of variations of are shown in Table 6. In this experiment, the control factors depends on orthogonal array L8 Taguchi design of experiment which was outlined in five factors and each factor had two levels. For the noise factors, the orthogonal array L1 Taguchi design of experiment was utilized as it had one factors and each had five intervels as summarized in Table 7.    A  B  C  D  1  1  1  1  1  2  1  2  2  2  3  1  3  3  3  4  2  1  2  3  5  2  2  3  1  6  2  3  1  2  7  3  1  3  2  8  3  2  1  3 To verify the impact each factor had on the yeild, the signal-to-noise (SN) ratio should have been ascertained for each experiment that was performed. The SN value denoting the mean of a process was compared to its variation. Three categories of SN ratio were figured in view of various kinds of execution attributes. In minimizing the attributes of the framework, the following SN ratio, which is called smaller-thebetter, was determined applying Equation 1, 2 and 3: Where y = mean response for experiment, i = number of experiment, u = number trial run, Ni = number of trials for experiment i. The third case is for nominal-the-best situation when a predetermined value is most preferred. To optimize the performance attributes, the following SN ratio, called the larger-the-better was anticipated as follows: For this analysis, the SN proportion larger-the-better was utilized for PDR and throughput assessment. For optimal performance, the larger-the-better performance metric for both PDR and throughput sensitivity was taken to acquire optimal VANETs congestion control plan for non-safety applications. Nonsafety applications that require low level quality of service (QoS), are delay sensitive. The investigation concentrated on non-safety applications in VANETs for highway driving situations. Table 8 clarified the execution measurements that were connected to highway experiment environments. Figure 2, abridges the stream and different stages of the trial procedure in the Taguchi optimization method for minimizing delay, maximizing PDR and maximizing the throughput for highway scenarios.  [32].
Packet Delivery Ratio (PDR) The number of packet received at the destination over the packet generated by the source [33].

Delay
The duration it takes for a data packets transmitted from source to destination [32].

RESULTS AND ANALYSIS
This segment shows the outcomes from the optimization design and simulation in OMNeT++. The delay performance attributes is represented in Figure 3 utilizing mean SN Plot for all reaction characters versus control factors. Figure 6 shown the average improvement propagation for delay sensitivity was 15.59% after optimization. At the emphasis purpose of 25KB and 50KB, the network speed expended, there was the likelihood that the data transmission suffered from latency at 50KB to 125KB. Other than that, AODV multi-hop and the point-to-multipoint data transmission of V2I likewise caused extra delay as a packet may need to wait for retransmitting of missing packets during transmission.
Nevertheless, this did not diminish the adequacy of transmission capacity utilization. The charts denoted the S/N ratio smaller is better which depended on the Taguchi method to acquire finest and smaller delay ratio giving it the best fit threshold setting. From the investigation, the optimal congestion control for vehicular system on reducing delay is appeared at 25KB to 50KB of packet size. The packet size demonstrated a little impact to congestion control measurement as expressed before where 15.59% was the average in terms of optimal packet transmission. This segment will introduce the outcomes from the optimal model in OMNeT++. The PDR and throughput performance attributes is outlined in Figure 4 and Figure 5 utilizing the mean SNR Plot for PDR and throughput response versus control factors. Figure 6 replicates the precision of the framework on the AODV protocol over congestion control towards multimedia applications after optimal packet transmission. Based on Figure 7, there is change after optimal process around 26.4% in viewed of packet delivery ratio (PDR) after optimal process, thus lowering packet loss. From 50KB onwards, the PDR is at a descending pattern that is reduce at a specific time since the modified AODV protocol increased latency of routing activity from the source to the destination. nodes. Based on the Taguchi method, the graphs in Figure 4, signifies the mean S/N larger is better in getting optimal and higher PDR transmissions with the best fit parameter setting. Figure 6 and Figure 7 demonstrate the mean SNR larger is better in the Taguchi investigation which is to acquire higher limit of throughput with respect to the best fit parameter setting. Based on Figure 8, when there was changed after optimization of 30.96% on the throughput, packet loss was additionally diminished after the optimization process. From 50KB onwards, the throughput is on upward pattern that is high at a specific time since the adjustment of AODV protocol generates less latency of routing movement from source to the destination. As a result the number of packets that could be broadcasted increased as well.

CONCLUSION
In this paper the authors examined innovations and techniques acknowledged by different researchers and proposed a system for congestion control for SCH applications focusing on non-safety applications utilizing the Taguchi optimization scheme. The congestion control approach is one of the better answers to ease congestion in a WLAN communications channel. The authors have featured the algorithm for the non-safety messages mechanism to lessen the channel communications utilization applying to the defined threshold.
Control factors and noise factors have indirect and direct impacts on occurrence of packet broadcasting in VANETs. A vigorous optimization technique is appropriate for various configuration factors, for examples, MAC protocols, routing protocols, networks topology and testbed environments. This paper proposed the Taguchi optimization method for improving the delay sensitivity, PDR sensitivity and throughput sensitivity. For future research trend, more attributes and conditions can be tested for congestion control.
There are various components that have indirect and direct effect on performances for non-safety or multimedia applications. These elements can be categories into twofold, control factors and noise factors. A vigorous optimization technique is appropriate for various configuration factors, for examples, MAC protocols, routing protocols, network topology and test environments. These experiments validated the Taguchi optimization method had enhanced the level of congestion control for multimedia applications in VANETs.
Packet size is a validated element in enhancing throughput and PDR. The distance of RSU also was a key factor in term of occurence in PDR. The simulation results demonstrate the AODV routing deployment has certain outcome in terms of packet transmission toward throughput and PDR. For future work, more parameters can be included for non-safety applications in the optimization process, for example number of nodes, number of vehicles, life time span and RSU distance.
As a conclusion, performance can be accomplished with a reduction in delay sensitive application for mobile user data traffic transmission that plans to maximize the overall effectiveness in transmission of non-safety packets. In this paper, the optimized technique that was setup analyzed the performance of the AODV routing deployment under two unique conditions which is when the changes have been applied in AODV routing in highway driving scenarios. Our model results demonstrate that the AODV routing deployment has constructive outcomes in terms of delay (number of broadcast packet received). The congestion control of the particular framework enhanced radically after optimization process for the vehicular ad hoc network at client or application level.