Impact Of Wlan On Bluetooth Video Streaming Computer Science

The advent of ubiquitous wireless technology makes it feasible to transmit real-time video in mobile and pervasive environments over Bluetooth. This paper elucidates the research and implementation of the factors affecting video streaming over Bluetooth RFCOMM protocol in the presence of WLAN in real experimental test scenarios. In our research we have implemented a test-bed comprising a Bluetooth video-streaming gateway (Server) and a Nokia mobile phone device (Client) to stream real-time video. The impact of the interference is investigated with respect to transmission time and data throughput.

KEYWORDS

Bluetooth, WLAN, RFCOMM, Video Streaming, Mobile, Multimedia

1. INTRODUCTION

Bluetooth is increasingly being used as the "last meter" technology for indoor universal wireless coverage. In such scenario, ad-hoc Bluetooth networks provides cost effective and simple communication of various mobile devices and are now targeted to extensively support multimedia traffic [1]. Two types of links Synchronous Connection Oriented (SCO) and Asynchronous Connectionless Link (ACL) are supported by Bluetooth .ACL links provide flexibility for transmitted data over network [4][2]. The Enhanced Data Rate (EDR) of IEEE 802.15.1 version 2.0 has a peak user payload of 2.2 Mb/s, which is the same as IPTV [4][2]. Bluetooth employees RF technology in the 2.4 GHz ISM band as 802.11b (WLAN). It uses frequency hopping spread spectrum (FHSS) technology at a rate of 1600 hops/s with a slot period of 625 milliseconds and hops over 79 MHz of bandwidth [6]. The WLAN uses Direct Sequence Spread Spectrum (DSSS) instead of FHSS. Its carrier does not hop or change frequency and remains centered on one channel that is 22 MHz-wide. While there is room for 11 overlapping channels in this 83 MHz-wide band, there is only room for three non-overlapping channels. Thus there can be no more than three different Wi-Fi networks operating in close proximity to one another.[6]

When a Bluetooth radio and a Wi-Fi radio are operating in the same area, the single 22 MHz-wide Wi-Fi channel occupies the same frequency space as 22 of the 79 Bluetooth channels which are 1 MHz wide. When a Bluetooth transmission occurs on a frequency that lies within the frequency space occupied by a simultaneous Wi-Fi transmission, some level of interference can occur, depending on the strength of each signal [4]. Streaming live video over wireless channel degrades the video quality and affects the data rates due to outside hindrance and co-channel interference. Our approach is the development of a test-bed implementing wireless connection between client (Mobile device ) and server ( PC ) and send data in the form of stream of byte array over RFCOMM (Radio Frequency oriented emulation of the serial COM ports on a PC) channel.

Past studies have showed that efficiency of Bluetooth is affected by wi-fi and Bluetooth interference [11]. According to the wave theory [12], when two waves overlap, they either superimpose constructively or destructively; moreover, when a wave passes through a surface, its intensity decreases greatly

Fig. 1 TEST SCENARIO

2. TECHNOLOGIES USED

J2ME / J2SE

Java 2 Platform, Micro Edition is a revolution in congruous to the diversities of the market aiming at embedded devices, cell phones, palmtop as new set of clients. It is prevailing due to its transformable, low cost and secure functional domain. J2ME is basically a subset of J2SE and most of its functionalities are derived from it [7]. The architecture of J2ME is divided into Configurations and Profiles. Configuration (CLDC/CDC) combines virtual machine (KVM for cell phones) is a set of predefined libraries showing device's power consumption, network connectivity and memory [8]. On the other hand profiles adds an additional layer of API's offering set of GUI's and local data storage. Java SE is a widely used platform for developing and deploying Java applications today's demanding embedded and real-time environments. It consists of a virtual machine, responsible for platform independence [3].

Fig 2 J2ME Architecture

Bluetooth

Bluetooth wireless technology is an open specification for low power, low cost, short range radio technology for ad-hoc wireless communication operating at 2.4 GHz and variable range of 10-100 m [1][9]. It uses radio technology called frequency-hopping spread spectrum, which divides the data being sent and transmits chunks of it up to 79 frequencies [6]. Bluetooth stack is the controlling agent for implementing the Bluetooth protocol and controlling Bluetooth device programmatically. The Bluetooth protocol is divided into Layers (Protocol Stack) and Profiles.

Fig 3 Bluetooth Protocol Stack

The essential steps for a Bluetooth communication includes: Stack initialisation, Device Management, Device Discovery, Service Discovery, Service registration, Communication [8][9]. The protocols used for the communication are: L2CAP (Handles data transmission in packets), RFCOMM (allows to create virtual serial port and to stream data), OBEX (allows sending and receiving objects) . RFCOMM forms a reliable peer to peer connection and is easy to implement. Similar to communicating over socket connection, data is sent via streams. L2CAP on the other hand supports multiplexing of devices and data is sent in packets. JSR 082 is used for its implementation in mobile devices using J2ME [7]

Fig 4 JSR 082 Taxonomy

3. IMPLEMENTATION FRAMEWORK

Our test-bed mainly consisted of Nokia (Symbian OS) S60 2nd Edition Feature Pack 3 mobile phone devices, specifically the Nokia N70 and Acer Laptop with Intel® Atom™ CPU N450 @1.66 GHz, 1.33 GHz, 0.99 GB of RAM and Windows platform. The choice of using Nokia is justified by its ubiquitous use in the market and wide availability of Symbian which supports variety of languages as C++, Java ME [1][13]. The implementation of the server (J2SE) and client (J2ME) modules were done using Netbeans IDE v 6.8. Microsoft's Bluetooth API supports the RFCOMM protocol, making it suitable and easier to use for our study. The test-bed consisted of two parts: Mobile device as a client and PC/Laptop as server. The architecture of the test-bed application has been illustrated in the figure above. The client side is developed using J2ME which makes it platform independent and to be used in various Java enabled devices. Server side is developed in Java SE using swings API and establishes a connection with client through Bluetooth. The protocol used is btspp [8] which transmits data in streams similar to socket connections. Only one channel exists between client and server for communication. Firstly, client searches for the Bluetooth server in the vicinity and requests for a connection with already ready server [13].

Fig 5 CLIENT SERVER ARCHITECTURE

After authentication, server establishes wireless connection with the client and displays the UUID of the client device on its interface and vice-versa. Secondly, the video/stills are captured by the client using MMAPI (Multimedia API) [7] and the player thread is started with the camera of the mobile device. The data is then taken and sent via channel in the form of stream of bytes to the server over RFCOMM connection. StreamConnection and StreamConnectionNotifier classes are used in the implementation of the test-bed. The video snapshots are sent continuously to the Bluetooth port and transmitted wirelessly to the server till the thread is stopped. The data is saved in the server folder and a simultaneous analysis of the size of data sent, time taken by the data to reach the server and rate at which the data is sent is examined. Different size data is sent across the server in range of different resolution and the size with maximum and consistent data rates is chosen for transmission. Results are simulated in the next section.

4. EFFECTS OF WLAN INTERFERENCE

After successful implementation of the framework with the hardware and software described, the effects of interference from Wi-Fi on transmission time and throughput are compared with situations where there are no interference while transmitting the video over RFCOMM protocol. As soon as a Bluetooth device encounters interference on a channel, it deals with the problem by hopping to the next channel and trying again [12]. Therefore, it attempts to avoid the interference from a WLAN network. When transmitting data on Asynchronous Connection-Less (ACL) links, the data throughput degrades substantially. On the other hand, when using Synchronous Connection Oriented (SCO) links to transmit time-sensitive information, packets can be lost because these links do not utilize Automatic Repeat Request (ARQ).

Figures 6 to 11 shows that interference has a negative effect on video transmission over Bluetooth and there is an increase in transmission time and a decrease in the throughput. It can further be noted that WLAN has the least effect for data size of 20 KB and the transmission rates and throughputs are constant. It is also noted that for small data size, WLAN increases the transmission by 2.5 seconds and affects overall transmission time linearly throughout. These sources of disturbances have a destructive impact on the wave frequency which in effect decreases the intensity [12] of the emission and thus increases the transmission time. As compared to Figure 8, WLAN has a greater impact on transmission for a data size of 200KB. The effect of Bluetooth interference is constant over the range of ten meters as shown by the constant horizontal line. The Wi-Fi interference affects the transmission time of data size of 200KB and increases it by 40%, Since the transmission time has been affected by the interferences, it is expected that the throughput will also be affected. This is shown in the graph above where wi-fi has the greatest impact on throughput. Moreover, Wi-Fi decreases the throughput by 21%.

It can further be noted that WLAN has the greatest impact on transmission for high data size. As previously stated, if transmission time is affected by interference and likewise throughput is affected. Figure 10 confirms this pattern.

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Fig. 10

Fig. 11

5. MINIMIZING THE AFFECTS

The coexistence of Bluetooth and Wi-Fi networks is defined as "the ability of one system to perform a task in a given shared environment where other systems may or may not be using the same set of rules". These practices fall into two categories[12]:

5.1 Collaborative mechanisms

The following collaborative techniques being considered require that a Wi-Fi device and a Bluetooth device be collocated (i.e. located in the same laptop).

TDMA (Time Division Multiple Access) : These techniques allow WLAN and Bluetooth to alternate transmissions. Bluetooth can support piconets, but cannot support SCO links.

MEHTA (the Hebrew word for "Conductor") : This technique is used for managing packet transmission requests. It grants permission to transmit a packet based on parameters including signal strength and the difference between 802.11 and Bluetooth center frequencies. It can support SCO links.

Deterministic frequency nulling: This mechanism is used in conjunction with MEHTA that inserts a 1 MHz-wide null in the 22 MHz-wide 802.11 carrier that coincides with the current Bluetooth center frequency

5.2 Non-collaborative mechanisms

A non-collaborative coexistence mechanism is one in which there is no method for the Bluetooth and WLAN to communicate. Non-collaborative techniques being investigated are:

Adaptive packet selection and scheduling is a Bluetooth Media Access Control (MAC) level enhancement that utilizes a frequency usage table to store statistics on channels that encounter interference. This table can subsequently be accessed by packet scheduling algorithms that schedule transmissions to occur only when a hop to a good channel is made.

Adaptive frequency hopping classifies channels and alters the regular hopping sequence to avoid channels with the most interference.

6. SUMMARY AND CONCLUSION

This paper elucidates the use of Bluetooth as a broadcasting medium for wireless transmission of data from mobile to PC. A test-bed was implemented successfully using Java and video streaming was done in real time. The impacts of interference due to WLAN on various data size on was investigated and studies with respect to change in data rates and time taken for transmission. The data was sent using RFCOMM protocol in the form of array of bytes from client to server. WLAN interference showed decrement in data rates and increase in transmission time because of the interference. It was also observed that the data transmission was dependent on data size. For small data sizes, the effect on transmission time was negligible and it became more significant as the data size increases. Distance from the Bluetooth server also affects the transmission time and throughput. As distance increases, transmission time increases while throughput decreases.

Article name: Impact Of Wlan On Bluetooth Video Streaming Computer Science essay, research paper, dissertation