TCOM 590, Fall 2013
Due on September 24th
Delivery method: hard or soft copy (email is just fine)
1. Assume audio conferencing session with multiple participants (RTP streams). How would you:
a. Identify deferent voice streams that belong to the same session? b. Differentiate between RTP and RTCP packets that belong to the same session? c. Conclude if there is as media mixer in this session?
d. How many voice milliseconds (ms) have been packetized if the RTP timestamp for the first packet, R1 = 980 and the RTP timestamp for the n-th packet Rn = 2100?
e. Indicate the beginning of talkspurt?
f. Reorder incoming RTP packets at the receiver end?
g. Determine the voice payload size if PCM is used to digitize voice signal? h. Detect RTP packet loss?
a. The SSRC in the RTP header helps identify voice streams that belongs to the same source in the session. The RTP SSRC helps label streams from different sources.
b. If the port number is even then the packet is an RTP packet. If the port number is odd then the packet is an RTCP packet. Usually RTP and RTCP(RTP+1) are assigned UDP ports from 16382-32500
c. Yes, there should and must be a media mixer in the conference session. The mixer essentially receives RTP streams from one or more sources, combines them in some way and forwards the new mixed stream to one or more receivers.
d. Timestamp to timevalue conversion : Rn-Ri /8000Hz
=) (2100-980)/8000 = 140µsec
Therefore, packets of size 140 µsec have been packetized.
The beginning of talkspurt is indicated by the marker bit. It also helps adjust the playout delay at the receiver and is also used for silence suppressions to indicate packets that follow periods of silence.
f. Sequence number in the RTP header helps reorder the RTP packets at the receiver end. And the timestamp helps play out the voice packets in sequence.
g. The payload type field in the RTP header indicates the payload type and size. If it’s PCM that’s used for digitization, then the payload type indicates the same. The sampling frequency being 64KHz in case of PCM the sample size or payload size will be 1/64K = 15.6 µsec per sample.
The sequence number in the RTP header helps detect the RTP packet loss
2. Consider voice conference session with 28 participants that each simultaneously sends and receives 102 kbps RTP streams. Determine RTCP packet transmission period for each sender and receiver assuming average RTCP packet size for the session to be 128 Bytes. Round your result to the nearest hundredths. (18 points)
# _ of _ senders
* avr. _ RTCP _ packet_ size
0.05* 0.25* session _ bandwidth
# _ of _ receivers
* avr. _ RTCP _ packet_ size
0.05* 0.75* session _ bandwidth
3. Draw and explain five common VoIP connection strategies we covered during the first lecture that transport voice over IP in at least one segment of the communication network. Make sure you include all required components to make interconnectivity between them possible. How VoIP systems communicate with each other and TDM legacy POTS systems. (18 points)
VoIP Connection Strategies
Strategy 1: VoIP Base to VoIP Base over IP Network
Strategy 2: VoIP Base to VoIP Base over Voice Network
Strategy 3: VoIP Base to POTS Base over IP Network
Strategy 4: VoIP Base to POTS Base over Voice Network
Strategy 5: POTS Base to POTS Base over IP Network
Strategy 6: POTS Base to POTS Base over PSTN
a. Compare following set of CODECs. Include algorithms they use and the efficiency of those, bandwidth requirements, delay impact, packet loos handling, VAD etc. In what specific situations would you use each of them? Explain your reasoning.
Most common today
A waveform codec
If uniform quantization were used, it would take 12...
Please join StudyMode to read the full document