Impact Significance And Contribution

Published Date: 02 Nov 2017

ABSTRACTS

This project proposed a program to adaptively encode the voice signal based on the available bandwidth detected in VoIP calls. It will provide a set of output sound with low voice data packets discard rate. Available bandwidth plays an important role in determine the quality of the output sound. "Dead Spot" or choppiness sounds, jitter, delay and packet loss can be considered as the symptoms of un-adequate available bandwidth used in VoIP calls. The program in this project uses JMF (Java Media Framework) as the software tools. RTP (Real-Time Transport Protocols) involve in the program implementation but will not advance discuss at this proposal. As SDLC (System Development Life Cycle) and Runtime.exec() act as the system development methodology, the program will be re-try and re-test until its meets the project objectives.

Chapter 1: Introduction

Motivation and Problem Statement

VoIP calls are getting familiar to public and it is useful for SME (small-medium Company) for business transaction due to cost effectiveness. However, some of the areas in our country still cannot provide a better bandwidth to support this. Public and SME need a better solution to meet with un-adequate network resources to overcome poor QoS in VoIP calls for business transaction. It is important for throughput estimation and encoder adaption project in VoIP calls to provide a preliminary support or solution to solve the problem.

VoIP Calls Overview

Before reviewing to the problem, there is a need to understand on how a VoIP calls works. To gain information on how the calls is establish, several research paper has been studied. Figure 1-1 show how a VoIP calls works:

Telephone

Personal Computer

Phone Adaptor

Cable Modem

Internet

Telephone

Personal Computer

Figure 1-1 VoIP Calls Flow

From the VoIP calls flow, the voice is converted ATA (Analog Telephone Adapter) from analog signal to digital signal and pass the signal through the Internet. VoIP can allow a computer, a special VoIP phone or a normal phone connected with a VoIP phone adaptor to communicate with each other. By the way, wireless hot-spot may allow computer to use wireless VoIP application to communicate with others. 

Problem Statement with VoIP

The main problem focus in this project for current VoIP calls is the bandwidth problem. VoIP needed high-bandwidth consumption to deliver high quality of voice signal. The available bandwidth used on VoIP calls can directly affect the voice quality. "Dead Spot" or choppiness sound occurs in VoIP calls is the symptom of un-adequate available bandwidth. This situation can happened in the incoming voice, outgoing voice or appear on both. Sometimes, voice especially outgoing voice will disappear for certain period of time. After a while, if the connection has not terminated by the caller/user, the voice appears again.

The low-bandwidth problems will causes packet loss, jitter and other issues. Latency is the time for the voice packet to arrive to its destination. Jitter is the variation of time between the arriving packets. Others issues like packet delay due to network congestion, timing drifts, or travelled route changes also critical to VoIP calls. Sorry to say that in VoIP calls, those are very serious matter as a small amount of time in packet loss, delay or jitter occurs will causes a poor voice communication.

Motivation

The motivation to develop a program to adaptively encode the voice signal base on the available bandwidth detected in VoIP calls is to help the SME with a better QoS for business communication. VoIP or IP telephony is a fast developing industry. Day by day, there is a new type of marketing services offer by the VoIP or IP telephony company. The growth of the VoIP today is almost the same as the growth of the Internet in 1990’s. The public can reap a lot of benefits at home or in their business and they started to pay attention in this industry.

With the help of advance technology, the QoS for voice signal in VoIP calls can be improved. There is a need in it for the entrepreneur with an unlimited long-distance call over state or country plus an acceptable standard of voice communication. Besides, there is also a need for a method to eliminate the problem of the unstable bandwidth internet connection. Since not everywhere can provide a stable bandwidth for the internet connection, it is a must for a method to solve the problems related to packet loss or choppiness sound (dead spot) in VoIP calls.

Project Scope

In the end of the project, a program to encode the voice signal base on the available bandwidth detected in VoIP calls will be introduced. This program will periodically detect the current available bandwidth and encode the voice signal which send the audio to another peer computer across the network. In the sender side, if the available bandwidth detected is in low value rtt or sufficient, the program will encode a higher or better quality of voice signal. If the bandwidth is in high value rtt or not adequate for a high quality voice signal, the program will encode a lower quality voice signal.

When the voice signal reach to the receiver side, "dead spot" or choppiness sound will not occur. The voice of the sender still can be heard even in low-bandwidth internet connection. In video, there is a high definition motion and standard definition motion. In this project, the voice in VoIP calls will be encode as the theory in better voice quality and standard voice quality.

Project Objectives

The main objective in this project is to increase the QoS (Quality of Service) in VoIP calls by efficiently used of the available bandwidth and adaptive encoder program. In the same time, by using the aid of advanced technology today, this project also can:

Able to detect the rtt value from another end device (computer).

Able to establish rtp session between two end devices (computer).

Provide encoded voice signal based on bandwidth detected from time to time in a VoIP calls.

Normally, packet will be drop in a call when the available bandwidth is not enough to supply the signal transmission. This project overcomes the packet drop with a lower quality of voice signal. Callers may experiences a short-period of weaker sound instead of choppiness sound when a lower available bandwidth is detected. The program focus on how is the voice signal is encoded based on the available bandwidth from time to time in a VoIP calls. Types of routing protocols used are not being focus in this project.

Impact, Significance and Contribution

This project is to improve the QoS in VoIP calls under unstable Internet Connection. Besides that, the program implemented can encode the voice signal based on the available bandwidth detected from time to time in a VoIP calls in order to eliminate choppiness sound problem.

With this program, the users can have a better visualization on how to detect the available bandwidth used and how to encode voice signal. It can act as a fundamental material to enhance the small-medium size IP telephony services. The users can experience the available bandwidth changes when make a one-to-one end devices VoIP calls. When there are changes in the available bandwidth, users can inspect how the voice signal is being encoded. This project has a lot of possibilities outcome to replace the regular phone services with a cheaper price and a better QoS.

Background Information

There are two fundamentals of technologies that are needed for the existence of VoIP. This first is the telephone, and the second is the Internet. The telephone was the works of Alexander Gram Bell and Elisha Gray in the 1870s. In 1968, the Internet is first developed by ARPANET (Advanced Research Projects Agency Network), founded by U.S. Department of Defence in 1957. ARPANET was developed to provide a decentralized communication network which will not be disrupted by the global war.

VoIP concept is to let the users to send data packets over Internet. This allowed PC users to avoid long distance charge and the first Internet Phone Software appear in 1995. VoIP evolved gradually in the next few years.  Although it is necessary to use a computer to establish it, VoIP start to involve in phone to phone services.

A step forward in VoIP history is when the hardware manufacturers starting to implement hardware which capable in switching.  It mean that what needed a computer to be done for "switching" a voice data packet to Internet is handled by the hardware and it results in making VoIP not so computer-based. Since now, VoIP usage has been increased and expands tremendously.  There are several different techniques in for voice data packet to transfer and switch among the network or Internet. Again, in this project, a program which adaptively encodes the voice data packet according to the bandwidth detected will be introduced.

Chapter 2: Literature Review

2.1. Literature Review

For the literature review, the sources will collect from the recent research paper relevant to bandwidth and encoder field. From the literature reviews, there are some useful research results which conducted by others IT researchers on throughput estimation and encoder adaption. Those research results can improve the concepts and processes on this project. Besides that, the discussion on how VoIP calls assist Small Medium Enterprise (SME) in their business operation will be cover here.

2.1.1. How VoIP calls assist Small Medium Enterprise (SME)

Most of the businesses owners today have switch their phone service to VoIP system phone service. The main reason is that VoIP system phone service is relatively cheaper than the normal standard phone company and it will help SME’s owners to save money. A gorgeous feature of VoIP technology is the affordability, due to the amount of software, manpower and hardware necessary to make VoIP service notably less expensive than the standard phone service.

The cost involved in VoIP system phone services can be different depend on the VoIP service providers offered. VoIP providers differentiate themselves from one another by offering even cheaper service with limited service. In another word, the cheaper the offer is, the more limitation on VoIP service the provider provides.

The benefits are:

Lower monthly telephone bills, including long distance calls.

Plans that have unlimited calls to a particular location.

Lower price in International calling.

Portability which mean can carry the phone service anywhere.

2.2. Facts Finding

According to Wojciech Mazurczyk1 and Zbigniew Kotulski (2007), the main problem in current VoIP services is the low perceived quality, which means the human evaluation on the poor voice quality they received in VoIP calls. Besides that, the services also do not provides good networking quality and utilize the available bandwidth usage. Poor networking quality may causes network condition like choppiness sound, jitter, delay or packet loss. Figure 2-1 shows the current problems in VoIP calls:

Internet

(Choppiness sound, jitter, delay or packet loss)

Receiver

Sender

Figure 2-1 Current Problems in VoIP calls

When a sender establishes a call, the analog signal is encode into digital signal and pass to the Internet. Inside the internet, there is a lot of routers need to travel in order to reach to the destination. At that time, different network topology, metric and others network issues would affect the signal and causes QoS degradation. When the digital signal encode back to analog signal in receiver side, this will affects the voice quality.

2.3. Data Collection

After reading some research papers and articles related to solution for poor quality in VoIP calls, there is some information needed to implement a solution. The information needed to implement the solution basically included the depth discussion of jitter sources, some JMF audio transmission examples, VoIP software tools.

2.4. Critical Remarks of Previous Works

2.4.1. Bandwidth Calculation for VoIP Networks Based on PSTN Statistical Model

Zvezdan Stojanovi´c and D- ord-e Babi´c shows an experiment on how to calculate bandwidth for VoIP calls after in proper PSTN network. For this, they use Erlang B and extended Erlang B which developed a software tool named Bandwidth Calculator. It can calculate the number of the circuits on PSTN and the IP bandwidth which required by the model. Their stimulation can be applied in any situation except PSTN. Their aim is to calculate the proper number of circuits for VoIP calls and Erlang B and Extended Erlang B models are used to determine proper number of circuits to carry voice traffic during the busy hour time (BHT). Their strength is they have proved that the value of Hurst parameter for dial-up traffic and for ADSL traffic is bigger than 0.7 and it is useful for them or other researches to stimulate bandwidth calculation on PSTN Statistical Model.

The limitation and weakness in their project are the stimulation they make in PSTN network to calculate the proper number of circuits for VoIP calls was not so practical. Environment factors changes across country areas and it will affect their project result which cannot be integrated. But, the limitation can be resolved by dynamically stimulate and estimate the proper number of circuits for VoIP calls area by area. In another word, one experiment in one typically area or town depends on the Internet access and network resources.

2.4.2. A Generic Technique for Voice over Internet Protocol (VoIP) Traffic Detection

Fauzia Idrees Uzma Aslam Khan mentioned about VoIP application and its generic technique. VoIP application like Skype, Google Talk, Yahoo voice and etc. offer cost effectiveness and are easy to use. Due to these reasons many new VoIP applications are coming into existence. A recent survey carried out predicted that VoIP will account for approximately 75% of world voice services by 2008 and the project state that the adoption of VoIP is complicated. Most of the identifying of the VoIP communications by different researches and commercial organizations are either specific to some protocol or router ports. Their strength is they had gained information regarding the characteristics of various VoIP applications and it is useful for them or other researches to implement a generic technique for VoIP traffic detection by those applications studied.

The limitation and weakness in their study are the VoIP packet is go through dynamic port and many protocols had appeared to support its implementation. Identifying the VoIP services accomplished through dynamic ports to avoid the traffic detection. But, the limitation can be resolved by provide a more detective mechanism in VoIP calls or services. Nevertheless, a new detection mechanism which is based on flow-level characteristics such as packet inter-arrival time, packet rate and packet-lengths had being introduced in this paper.

2.4.3. Adaptive VoIP with Audio Watermarking for Improved Call Quality and Security

Wojciech Mazurczyk1, Zbigniew Kotulski describe a novel adaptive method of speech quality control which may be used to adjust three call parameters: speech codec configuration, play-out buffer size, and amount of FEC (Forward Error Correction) mechanism information during VoIP (Voice over Internet Protocol) call under changing network conditions and it is a bit similar with my projects concepts which is also improved call quality. This is because it utilizes audio watermarking techniques as a communication channel between calling parties to send information and change in quality of the call. This paper states that VoIP is very popular and plays an important role in the telecom market which is real-time service that enables conversation through IP networks. However, there are two unsolved problems still exist for IP telephony. One of them is providing security of the traffic between calling parties and the other is providing practical quality of the call for end-users. It is important for the second issue because caller need some standard of conversation calls over VoIP to be able to communicate among each other. Their strength is they have found an effective solution for adaptive VoIP application and it is useful for them or other researches to improve the call quality and security in VoIP with Audio Watermarking.

The limitation and weakness in their project are DRM (Digital Right Management) involve and speech quality versus security mechanism. Although the primary application of audio watermarking was to preserve copyright and/or intellectual properties, but it become more complicated and expensive in the implementation of VOIP calls and the involvement of DRM cause a lot of telephony companies a pain in the neck due to the increment of service costs. But, the limitation can be resolved by combining speech quality control and providing security into one solution. In another word, they are combining 2 services into one mechanism.

2.4.4. Bandwidth Estimation Algorithms for the Dynamic Adaptation of Voice Codec

Davide Pierattoni, Ivan Macor, and Pier Luca Montessoro mentioned about new algorithm to control the resource utilization and to optimize the voice codec selection during SIP (Session Initiation Protocol) call setup on behalf of the traffic condition estimated on the network path. This project state that most of the connection-oriented communications based on the TCP protocol are able to react during the network congestion and most of the estimation bandwidth over a network are implemented in experimental tools. The use of a codec with high compression is the most simple and practicable solution for controlling and reducing the bandwidth occupation of a VoIP application. Their strength is they had gained experiment results of their bandwidth estimation algorithm under varying network conditions and it is useful for them and other researches to proceed in better dynamic adaptation of Voice Codec in SIP.

The limitation and weakness in their study are the estimation of capacity and available bandwidth relies on efficient technique and it is varies among country and protocols. Traffic overhead and its measurement on different network paths in voice codec also are the main problems in their study. But, the limitation can be resolved by using their algorithms in determine the best codec on the network paths. Nevertheless, it is notable that the end-to-end SIP experiments do not require any feedback from the network so that the estimation of capacity and available bandwidth problems can be overcome by their algorithm.

2.4.5. Comparison of Existing Project and Proposed Project

After reviewing the existing research project papers, there have own strength and weakness. The proposed project will occupy some of the strength and eliminate some of the weakness in the existing proposed project. Table 2-1 shows Comparison between the existing project and proposed project:

Bandwidth detectable

Adaptive Signal Encode

Protocols / Compatible with

Encryption Involved

Bandwidth Calculation for VoIP Networks Based on PSTN Statistical Model

No

No

RTP, PPP, IP, UDP

Not stated

A Generic Technique for Voice over Internet Protocol (VoIP) Traffic Detection

Yes

No

H. 323

Yes

Adaptive VoIP with Audio Watermarking for Improved Call Quality and Security

No

Yes

RTP, RTCP

No

Bandwidth Estimation Algorithms for the Dynamic Adaptation of Voice Codec

No

Yes

SIP, TCP

No

Proposed Project: Throughput Estimation and Encoder Adaption in VoIP calls

Yes

Yes

RTP

No

Table 2-1 Existing and proposed project Comparison

Chapter 3: Methodology

3.1. Methodology and Tools

3.1.1. System Development Life Cycle (SDLC)

The methodology for the program used in this project is the System Development Life Cycle (SDLC). Figure 3-1 shows the development process of SDLC:

Planning

Analysis

Design

Implementation

Testing

Figure 3-1 System Development Life Cycle (SDLC)

First, need to define what the aims and objectives are for this project. Plan the overall program flow and scenario occurred. Second, analyze the problem occurred and what can be done to modify and enhance the current VoIP calls. Defining whether the standard of the prototype in throughput estimation and encoder adaption program is correct with the project scope and objectives. Third, specify the programming language, software tools and applications needed in this project. Program flow design also needed to design until the experiment reach adequate standard. Coding will be started in implementation phase. This phase will be proceeding in Final Year Project II and this project still has space for innovation and flexibility. Finally, the program will be tested whether it can detect bandwidth in current network and whether it can encode the signal adaptively. Result will be recorded down and re-test if it do not reach the standard which had being defined in Analysis phase.

3.1.2. JMF (Java Media Framework)

The Java Media Framework (JMF) handles time-based media, media which changes with respect to time and is a recent API for Java dealing with real-time multimedia presentation and effects processing.  JMF will be used in this project to establish audio connection.

3.1.3 RTP (Real-time Transport Protocol)

RTP was designed by IETF’s (Internet Engineering Task Force) Audio-Video Transport Working Group to support video or audio conferences and it is an Internet protocol standard to manage the real-time transmission of multimedia data over network services. RTP does not provide guarantee end-to-end services itself and it is commonly used in Internet telephony applications. In this project, RTP involve to transfer audio packets.

3.1.4 Articles, Journals and Research Papers

In this project, articles, journals and existing research papers in VoIP field plays an important role. Useful information and documents to justify the problem statement and why there is a need for a solution on it about VoIP calls could be obtained. Methodology and tools to implement the program in this project can be decided by study the existing research papers.

3.2. Implementation Issues and Challenges

There are several implementation issues in this project. If the program fails to detect the available bandwidth within a single VoIP calls, it fails to meet with the objectives which have been initially defined. When implement the program in this project, it is a must to ensure that the throughput estimation part is error-free.

Since there is a rapid development in the VoIP software field, it is unconfident that the performance output of voice quality through this program would practically as same as the output when occupy it in the latest VoIP software. Besides, the performance output of voice quality may differ when making one-to-one end devices VoIP calls in the outside world compared to the experiment implementation.

3.4. Requirement Specification

The requirement specification is the overall specify details about the program which will be implemented in this project. Basic requirements details will be discussed here.

3.4.1. User Requirement

Typically the target user for this program is someone who has a basic knowledge in data communication field. The main purpose of this program is to carry out reliable voice quality data to improve the current VoIP services shortage. Mainly this program alert user with 2 sections, the throughput estimation parts and the encode parts. For the throughput estimation parts, user will know whether the current available bandwidth using is adequate or not. For encode parts, user will heard different quality of sounds depend on the bandwidth detected.

3.4.2. System Performance Definition

To make sure the program will run smoothly, the combination in software and hardware requirement is needed to take into consideration. Besides that, different kind of network condition needed to be adjusted and identify to practically virtualize the real-world network situation.

Below is the sample of the hardware and software needed in this program:

Computers with Intel I 3 processor with 2.75GHz processing power and above

Computers with 4GB DDR3-1333MHz SDRAM

JMF tools

Wireshark

Stable or unstable network connectivity

Chapter 4: Evaluation

4.1 Design Verification Plan

There are some steps which need to be done in order to implement the project. The first thing is to install the JDK (Java Development Kit), JRE (Java Runtime Environment) and JMF 2.1.1e. Make the jar files in JMF 2.1.1e into library and import in your development tool. In this project, NetBeans IDE 7.2.1 will be the development software tool. This project need two computer which connect to the same network devices which mean among the same local area network. It should able to ping among each other and get the rtt value. The AudioFormat use to encode the sound capture and define which audio format to be used for the low or high sound quality. Finally, the encoded sound will be received by JMF or a simple receiver program on the receiver side. Figure 3-2 shows the necessary steps to implement this project.

Figure 3-2 System Implementation Steps

4.2 Testing scope and Effort

Figure 4-1 will shows the effort used and the scope in the program testing.

Figure 4-1 Testing Scope and Effort

It will be something like step-by-step testing as mentioned in 4.1 Design Verification Plan. The first testing is to test that whether the two end devices can be connected within a same local area network which is the easiest part. Then check whether the throughput estimation which is a ping program works or not. Next, check whether the sound can be pass through the network and receive by the receiver. The last part and the most difficult part is to check whether the sound format can be changed based on the rtt time detected.

Chapter 5: Conclusion & Future Work

In this project, a program to encode the voice signal based on available bandwidth detected has being proposed which may enhanced VoIP calls in perceived quality for end-users. It able to detect the rtt value from another end device, able to establish rtp session between two end devices and /*can provide encoded voice signal based on bandwidth detected from time to time in a VoIP calls.*/ {Still trying} For my own experience, the project teach me a lot about JMF and Java programming to establish and perform rtp connection. What frustrated me is when I deal with the AudioFormat parts. Maybe I not good in programming because there around fifty-times I try to change the AudioFormat but it’s failed. Thanks to my supervisor, Dr. Saleem for giving me a lot of guidelines and hints to make this project done. Really appreciate his patient and teaching.

Since most of the parts in this program skeleton works, it can be advanced to become an audio/ video conferencing program. The quality of the audio/ video can be adjusted manually or adaptively while people doing audio/ video conferencing with each other. By the way, it may help students or someone who is interested in JMF as a foundation program to be learned, modified and improved.