240 likes | 467 Views
NETW 707 Modeling and Simulation Amr El Mougy Maggie Mashaly. Lecture (6) Traffic Modeling and Simulation. Why is it Needed?. Router design relies heavily on traffic modeling Quality of service support can be significantly improved if the traffic can be predicted
E N D
NETW 707 Modeling and Simulation Amr El Mougy Maggie Mashaly
Lecture(6) Traffic Modeling and Simulation
Why is it Needed? • Router design relies heavily on traffic modeling • Quality of service support can be significantly improved if the traffic can be predicted • Congestion control can be optimized by learning about traffic • Realistic simulations need realistic traffic Traffic intensity La/R < 1
Traffic Modeling • Objective: to simulate network traffic • What is the purpose of the simulation? Which layer is of interest Application Transport Network Data Link PHY
Traffic at the PHY Layer 01000110101010101011000100111100101011 • Sequence of bits • The bits are either there or not (ON/OFF) • Simulations at the PHY layer are usually concerned with BER and channel quality • Often simulations at the PHY layer assume constant stream of bits
Traffic at the MAC Layer • Starting at the MAC layer, packets are seen as black boxes • If medium access is to be investigated, then interference may need to be modeled • What is the MAC protocol used (ACKs, NACKs, GoBackN, Selective Repeat, etc.) • Is random access employed (are collisions possible)
Traffic at the Network Layer • IP traffic can also be modeled as an ON/OFF process • What is the routing protocol used
initiate TCP connection RTT request file time to transmit file RTT file received time time Traffic at the Transport Layer • A packet is usually followed by an ACK in the other direction, typically after one half RTT
peer-peer client/server Traffic at the Application Layer • Application layer protocols have different characteristics • Client-server or peer-to-peer architectures
HTTP • Web browsing • Uses TCP • persistent or non-persistent connections request line (GET, POST, HEAD commands) status line (protocol, status code, status phrase) HTTP/1.1 200 OK Connection close Date: Thu, 06 Aug 1998 12:00:15 GMT Server: Apache/1.3.0 (Unix) Last-Modified: Mon, 22 Jun 1998 …... Content-Length: 6821 Content-Type: text/html data datadatadatadata ... GET /somedir/page.html HTTP/1.1 Host: www.someschool.edu User-agent: Mozilla/4.0 Connection: close Accept-language:fr (extra carriage return, line feed) header lines header lines Carriage return, line feed indicates end of message data, e.g., requested HTML file
SMTP • SMTP uses TCP connections S: 220 smtp.example.com C: HELO relay.example.org S: 250 Hello relay.example.org, I am glad to meet you C: MAIL FROM:<bob@example.org> S: 250 Ok C: RCPT TO:<alice@example.com> S: 250 Ok C: RCPT TO:<theboss@example.com> S: 250 Ok C: DATA S: 354 End data with <CR><LF>.<CR><LF> C: From: "Bob Example" bob@example.org C: To: "Alice Example" <alice@example.com> C: Cc: theboss@example.com C: Date: Tue, 15 January 2008 16:02:43 -0500 C: Subject: Test message C: C: Hello Alice. C: This is a test message with 5 header fields and 4 lines in the message body. C: Your friend, C: Bob C: . S: 250 Ok: queued as 12345 C: QUIT S: 221 Bye {The server closes the connection}
Other Applications YouTube • Uses HTTP for pages and RTMP for streaming videos • May use TCP or UDP • Can be an example of Constant Bit Rate (CBR) traffic • Skype • Uses TCP and UDP • Proprietary protocols • Can be an example of Variable Bit Rate (VBR) traffic
Parameters for Traffic Modeling • Two parameters are needed to model traffic of any type • Packet size • Inter-arrival times • Packet size is easy to model. May be subject to protocol restrictions • Inter-arrival times are more challenging
Classic Traffic Modeling: The Poisson Distribution • Used originally to model arrivals of calls in a telephone network • The distribution is memoryless. Assumes arrivals are independent • Very simple and easy to use • Can be seen as a counting process. Inter-arrival times follow an exponential distribution • Has a single parameter, λ • Mean and variance also equal to λ
Properties of the Poisson Distribution The superposition of independent Poisson processes results in a new Poisson process with a rate equal to the sum of the rates of the independent processes Aggregate Arrivals
Trouble with Poisson Does not show traffic burstiness over extended time scales
Compound Poisson Traffic • The model is extended to deliverbatchesof traffic at once. • The inter-batch arrival times are exponentially distributed, while the batch sizes are geometric. • The model has two parameters: • The mean inter-batch arrival time 1/λ • The batch parameters ρ (between 0 and 1) • Thus, mean packet arrival over time period t is tλ/ρ • Disadvantages: • Back to back packet arrivals may not be realistic (although now it is more likely) • The model is still essentially Poisson, which is memoryless
Markov Modulated Poisson Traffic Model • Motivated by the need to generate packet arrivals at different rates • A continuous-time Markov chain varies the arrival rate of a Poisson model • Each state in the Markov chain has an associated arrival rate • For example, a two state MC has four parameters (λ1, λ2, r1, r2). • To determine these parameters, real traffic traces must be used • The model is designed to fit the real trace based on metrics such as: mean packet arrival rate, variance-to-mean ratio of the number of arrivals over a short period, or long-term variance-to-mean ratio of the number of arrivals r1 λ1 λ2 1-r2 1-r1 r2
The Packet Train Model • Recognizes the fact that address locality applies to routing decisions, i.e. packets generated by the same source with small inter-arrival times are probably bound to the same destination and thus will probably follow the same route • Packet trains are characterized by tandem trailers: a group of packets going in one direction, followed by one or more packets in the other direction • Characterized by 4 parameters: inter-train arrival time, inter-car arrival time, mean train size, mean car size • Does not make any decisions about the protocols and their nature
ON/OFF and the Interrupted Poisson Process • Two-state systems used to model the channel • Packet arrivals occur during the ON state according to a Poisson distribution • The time the channel spends in each state is called the transition time