Simulation Parameters

The simulation model for the real-time video stream with CBR and VBR background traffic is considered for stationary nodes and mobile nodes as shown in Table 5-1. Both the stationary and mobile nodes simulation model consists of four wireless nodes over a network simulation area of 500m X 500m. The wireless propagation model in Section 3.4.5 is applied. The overall network architecture is shown in Figure 5-1. The density of nodes will have an effect on the performance of the network, however four nodes have been chosen because it is close to the minimum needed to demonstrate the performance of the scheduler with realistic traffic.

Node 1 is the source node; node 2 and node 3 are the intermediate nodes whilst node 4 is the destination node. The video stream as well as the background traffic is sent from the source (node 1) to the destination (node 4). A Logitech webcam is connected to the computer running OPNET. Node 1 (source) captures the video stream from the webcam connected to the PC; it sends the streams to node 4 (destination) via the intermediate nodes 2 and 3. The video stream can be viewed at the destination node. The video caption, as viewed from the source and destination, is shown in Figure 5-2.

Oche Alexander Egaji Page 117 Node 3 Node 4 Node 1 Node 2 Camera Video Display

Figure 5-1: Video Streaming Model

Figure 5-2: Video display for Source and Destination

The interface between OPNET and the webcam is created using the Open source Computer Vision library (OpenCV), a C programming library. The implementation is shown in Appendix G. The highlighted parts of the codes are the changes made to the standard model to accommodate the video streaming in the node 1 and video display at node 4. The captured video frame format is not supported by OPNET. Thus, a custom packet is created within the OPNET simulation model to encapsulate each captured frame from the camera before transmitting it through the network. The packet is sent from node 1 to node 4 via intermediate nodes (node 2 and node 3). At node 4, the packet is de-capsulated, and the video can be viewed.

Oche Alexander Egaji Page 118 The resolution of the video frame is 160 x 120 pixels. This is converted to bytes to give the required payload size. Each pixel in a coloured image has three channels (red, green and blue), these colours are represented by 1-byte. Thus, individual pixels in a coloured image consist of 3-byte. A grey scale image consists of a single channel, which is 1-byte. Each byte consists of 8-bits; hence, a 160 x 120 pixels grey scale image would occupy a size of 160 x 120 x 1 bytes, which is 19,200 bytes or 153,600 bits. For simplicity, the video stream is converted to grey scale before being sent through the network. The maximum allowable packet size or Maximum Transmission Unit (MTU) that can be transmitted through OPNET network is 2,304 bytes i.e. 18,432 bits. This is the maximum frame body length defined in IEEE 802.11 standard [109]. Packets greater than the maximum allowable size are fragmented before being sent through the network, and it is re-assembled at the destination node before being displayed. The simulation parameters for stationary and mobile nodes are shown in Table 5-1.

Table 5-1: Simulation Parameters for Stationary and Mobile Nodes No. of Nodes 4

Area 500m x 500m Simulation Time 600s

Mobility Model Stationary Nodes-None

Mobile Nodes – Random Waypoint Propagation model Shadowing model

Data payload 153,600bits / 19,200bytes Frame rate 10fps

Traffic intensity ‘𝜌’ 0.33, 0.45, 0.60, 0.75, 0.90, 1.05, 1.55 and 2.05 MAC protocol IEEE 802.11n (Buffer Size = 16MB)

The OPNET network topology for stationary and mobile nodes is shown in Figure 5-3 and Figure 5-4, respectively. The OPNET node model ‘MANET station fixed’ is used for the stationary nodes whilst the ‘MANET station mobile’ is used for the mobile nodes. According to Figure 5-3, Node_1 is the source node, it transmit the video stream and background traffic; Node_2 and Node_3 are the intermediate nodes, whilst Node_4 is the destination node for the video and background traffic. This is similar to Figure 5-4, where Mobile_1 transmits

Oche Alexander Egaji Page 119 the video stream and background traffic, Mobile_2 and Mobile_3 are the intermediate nodes, whilst Mobile_4 is the destination.

Figure 5-3: Stationary nodes (Ad-hoc network)

Figure 5-4: Mobile nodes (MANET)

The source nodes transmit video streams at a rate of 10fps, with a frame size of 153,600 bits, the traffic intensity ‘



’ according to section 3.2.2 with no background load is 0.33;

Oche Alexander Egaji Page 120 this will form the baseline upon which other higher traffic intensity will be compared. An example hardware device that can interface the simulation network is the DRK8080. The DRK8080 is an integrated high bandwidth WIFI 802.11 Robot that supports video streaming, audio and sensors information. The system can upload the sensor as well as stream video (up to 4fps) to home PC [110]. The traffic intensity



_{is increased gradually} from the baseline (



=0.33) by introducing background traffic, and the performance of the video stream when FIFO, Manoj and both proposed schedulers are applied will be analysed. The different level of background traffic introduced as shown by the traffic intensity



in Table 5-1 represents 0, 26, 44, 56, 63, 68, 78 and 84% respectively of the total traffic in the network. When the traffic intensity



is greater than or equal to 1, the network is congested. The proposed Mamdani and Sugeno schedulers assess individual packets and assign the highest Priority Index to the video traffic because of their sensitivity to delay and loss. The previous simulation in Chapter 4 did not consider a network with mixed traffic. The priority of the background traffic is determined as explained in section 4.3.1.

In document Quality of Service for Multimedia and Control System Applications in Mobile Ad-hoc Network (Page 136-140)