Advanced Computer Networks cs538, Fall 2014 @ UIUC Klara Nahrstedt Lecture 8, September 18, 2014 Based on Zheng Wang and Jon Crowcroft, “QoS Routing for Supporting Multimedia Applications”, IEEE JSAC 1996

Announcements Next Reading for Thursday, September 26: X. Yang, D. Clark, A. Berger, “NIRA: A New Inter-Domain Routing Architecture”, IEEE/ACM Transactions on Networking, Vol. 15, No. 4, August 2007 Proposals for Final Projects Prepare project proposal At most 1 page describing Problem description Steps you plan to take to address the problem Related work (at least 3 full academic papers citations) and why your proposed problem is different than those or why your proposed solution is better. Deadline for proposal: 11:59pm Tuesday, September 23, 2014 No CLASS on Tuesday, September 23, 2014 Submit project proposal via email to instructor: klara@Illinois.edu (with subject: cs538 – Final Project Proposal)

Outline QoS Concept Single Mixed Metric Multiple Metrics Bandwidth and Delay QoS Routing Problem Path Computation Algorithms Source Routing Algorithms Hob-by-Hop Routing Algorithms

Window of Resources 1980 1990 2000 2010 2020 Requirements Interactive Network File access High-quality Audio HDTV Interactive HDTV-quality multi-view video abundant Sufficient Sufficient but Scarce to Insufficient - But scarce 1980 1990 2000 insufficient 2010 2020 To sufficient insufficient abundant Hardware support CS 414 - Spring 2011

Quality of Service (How to parameterize services?) To manage resources, we need services over resources to schedule AV data, to shape access for AV data, to process AV data, to move AV data, etc. Multimedia systems consist of set of AV-specific services Processing (media-related) services: retrieve audio/video, record video/audio, compress audio/video, fast forward video, rewind video Transport (network) services: Stream video, fast forward video, rewind video To provide multimedia services, services get parameterized with quality levels called Quality of Service CS 414 - Spring 2011

Layered Model for QoS Quality of Experience Quality of Service CS 414 - Spring 2011

Application AV QoS Parameters QoS for Audio service: Sample rate – 8000 samples/second (8KHz), 44.1 KHz Sample resolution – 8 bits per sample, 16 bits per sample QoS for Video service: Video frame rate – 25 frames per second, 30 frames per second Frame Period – 40 ms, 30 ms, 25 ms, … Frame resolution – 320x240 pixels, 640x480 pixels, 1920x1080 pixels, … Pixel resolution – 24 bits per pixel, 8 bits per pixel Frame size – 64KB Compression rate – 8:1 CS 414 - Spring 2014

Network QoS Bandwidth – Rate of data transfer, Bit Rate e.g., 1 Gbps (Ethernet throughput) – level 1 e.g., 100 Mbps (WiFi throughput) – level 2 e.g., 128 kbps (ISDN throughput) – level 3 measured in bits per second Throughput – rate of successful message delivery over communication channel Measured in packets per second, data packets per time slot, or bits per second 30 packets per second; 128 kbps, 10 packets per time slot CS 414 - Spring 2014

Network QoS Connection setup time Error Rate time how long it take to connect the sender and receiver e.g., 50 ms, 10 ms, … Error Rate Measures the total number of bits (packets) that were corrupted or incorrectly received compared with the total number of transmitted bits (packets) Bit Error Rate (BER) – at physical/MAC layer In fiber optics, bit error rate (BER) is of the order of 10-8 to 10-12. In satellite networks, BER is of the order 10-7 Packet Error Rate (PER) – at IP/transport/application layer – also called Packet Loss Rate CS 414 - Spring 2014

Network QoS Delay End-to-End Delay Latency Response time End-to-end delay in telecommunication Response time Round-trip delay in telecommunication End-to-End Delay time interval from the time packet is sent from the sender until the time it is received at the receiver (Treceive – Tsend) e.g., 80 ms, 100 ms, 160 ms CS 414 - Spring 2014

Network QoS Response Time Measured as round-trip delay and is the total time required for sender to send a packet and receive an acknowledgement from the receiver. It can be described as sum of network delay and interface delay. Network delay – composed of transit delay and transmission delay Transit delay is caused by time needed to send data on a physical connection between sender and receiver Transmission delay is time needed to transmit packet through network as result of processing delays (e.g., look up routing tables) Interface delay – incurred between the time a sender is ready to begin sending and the time a network is ready to accept and transmit the data (due to traffic policing and shaping) CS 414 - Spring 2014

Other QoS Parameters Jitter Undesired deviation from true periodicity in telecommunication Also called packet delay variation – important QoS factor in assessment of network performance Packet jitter – variation in latency as measured in the variability over time of the packet latency across network. CS 414 - Spring 2014

QoS Classes Guaranteed Service Class Predictive Service Class QoS guarantees are provided based on deterministic and statistical QoS parameters Predictive Service Class QoS parameter values are estimated and based on the past behavior of the service Best Effort Service Class There are no guarantees or only partial guarantees are provided CS 414 - Spring 2011

QoS Classes (cont.) QoS Class determines: (a) reliability of offered QoS, (b) utilization of resources CS 414 - Spring 2011

Deterministic QoS Parameters Single Value: QoS1 – average (QoSave), contractual value, threshold value, target value Throughput – 10 Mbps Pair Value: <QoS1, QoS2> with QoS1 – required value; QoS2 – desired value <QoSavg,QoSpeak>; <QoSmin, QoSmax> Throughput - <8,12> Mbps CS 414 - Spring 2011

Deterministic QoS Parameter Values Triple of Values <QoS1, QoS2, QoS3> QoS1 – best value QoS2 – average value QoS3 – worst value Example: <QoSpeak, QoSavg, QoSmin>, where QoS is network bandwidth Throughput <12, 10, 8> Mbps CS 414 - Spring 2011

Guaranteed QoS We need to provide 100% guarantees for QoS values (hard guarantees) or very close to 100% (soft guarantees) Current QoS calculation and resource allocation are based on: Hard upper bounds for imposed workloads Worst case assumptions about system behavior Advantages: QoS guarantees are satisfied even in the worst case case (high reliability in guarantees) Disadvantage: Over-reservation of resources, hence needless rejection of requests CS 414 - Spring 2011

Predictive QoS Parameters We utilize QoS values (QoS1, ..QoSi) and compute average QoSbound step at K>i is QoSK = 1/i*∑jQoSj We utilize QoS values (QoS1, , QoSi) and compute maximum value QoSK = max j=1,…i (QoSj) We utilize QoS values (QoS1, , QoSi) and compute minimum value QoSK = min j=1,…i (QoSj) CS 414 - Spring 2011

Best Effort QoS No QoS bounds or possible very weak QoS bounds Advantages: resource capacities can be statistically multiplexed, hence more processing requests can be granted Disadvantages: QoS may be temporally violated CS 414 - Spring 2011

Quality-of-Service Routing Audio/Video Multimedia Applications Real-time requirement Throughput requirement Sustainable performance Routing: A Key Network Function to Support QoS Diverse QoS constraints (NP-complete problems) Best-effort traffic and QoS traffic Dynamic network state

QoS Routing – Single Metric Bandwidth 50Mbps 30Mbps 40 Mbps 100 Mbps 50 Mbps 60 Mbps 120 Mbps S D Minimum Metric b(S,D) = min(b1, b2, b3, …, bn)

QoS Routing – Single Metric Delay 30 ms 40 ms 50 ms 15 ms 60 ms 120 ms S D 100 ms Additive Metric d (S,D) = d1 + d2 + …. dn

QoS Routing with Single Mixed Metric (50 Mbps, 30 ms, 0.02) (30 Mbps, 35 ms, 0.5) (40 Mbps, 40 ms, 0.4) (100 Mbps, 20 ms, 0.01) (50 Mbps, 35 ms, 0.1) (30 Mbps, 40 ms, 0.2) (120 Mbp, 15 ms, 0.01) (50 Mbps, 30 ms, 0.2) (60 Mbps, 25 ms, 01) S D F(p) = B(P)/ (D(p) x L(p), where B is bandwidth, D is delay, L is loss probability for path p; path with large value is

Multiple Metrics Problem: Find path subject to multiple constraints This is NP-complete problem Simple problem with two constraints is called “shortest weight-constrained path” Definition: d(i,j) be a metric for link (i,j). For any path p = (i,j,….l,m), we say metric d is additive if d(p) = d(i,j) + d(j,k) +…+ d(l,m). Metric d is multiplicative if d(p) = d(i,j) x d(j,k) x … x d(l,m) Metric d is concave if d(p) = min[d(i,j), d(j,k), …, d(l,m)

Path Computation Algorithms – Source Routing – Bandwidth-Delay Routing Consider directed graph G=(N, A) with N number of nodes (vertices) and A arcs (edges). (i,j) is arc (edge) bij be available bandwidth dij be propagation delay p = (i,j,k,…, l,m) Bottleneck bandwidth of path is width (p) = min(bij, bjk,…., blm) Length of path is length (p) = dij + djk + … + dlm Given any two nodes i and m of the graph and two constraints W and D, the QoS routing problem is then to find a path p* between i and m so that width (p*) ≥𝐵 and length(p*) ≤𝐷. Bandwidth-delay constrained paths

Routing Algorithm S D Constraints: B = 50 Mbps, D = 55ms (50 Mbps, 30 ms) (60 Mbps, 35 ms) (40 Mbps, 40 ms) (100 Mbps, 20 ms) (50 Mbps, 35 ms) (50 Mbps, 40 ms) (120 Mbp, 15 ms) (50 Mbps, 30 ms) (60 Mbps, 25 ms) S D Constraints: B = 50 Mbps, D = 55ms Step1: find all paths that satisfy B Step 2: find the final path that satisfies D

Hop-by-Hop Routing (50 Mbps, 30 ms) (60 Mbps, 35 ms) (40 Mbps, 40 ms) (100 Mbps, 20 ms) (50 Mbps, 35 ms) (50 Mbps, 40 ms) (120 Mbp, 15 ms) (50 Mbps, 30 ms) (60 Mbps, 25 ms) S D With multiple metrics the best path with all parameters at their optimal values may not exist at all!!!!

Hop-by-Hop Algorithm (1) (50 Mbps, 30 ms) (60 Mbps, 35 ms) (40 Mbps, 40 ms) (100 Mbps, 20 ms) (50 Mbps, 35 ms) (50 Mbps, 40 ms) (120 Mbp, 15 ms) (50 Mbps, 30 ms) (60 Mbps, 25 ms) S D Shortest-widest algorithm with distance vector approach Step 1: find all widest path from node 1 to each node ‘i’. If there are more than one widest path found, Step 2 chooses one with minimum length Step 3 updates the width and length for the shortest-widest path from node 1 to node i (using distance vector approach)

Hop-by-Hop Algorithm (2) (50 Mbps, 30 ms) (60 Mbps, 35 ms) (40 Mbps, 40 ms) (100 Mbps, 20 ms) (50 Mbps, 35 ms) (50 Mbps, 40 ms) (120 Mbp, 15 ms) (50 Mbps, 30 ms) (60 Mbps, 25 ms) S D Shortest-widest algorithm based on link state Step 1: find the nodes with maximum width among the tentatively labeled nodes if there are more than one node found Step 2 chooses one with minimum length and permanently labels Step 3 updates the tentatively labeled nodes around the new permanently labeled node

Conclusion QoS Routing is integral part of resource management QoS routing might be integrated with path reservation Status QoS has been explored in routers, but not much used QoS has been used in ATM networks (backbone) QoS service class concept is now being explored by broadband providers for multimedia services QoS in wireless challenging and statistical over unlicensed spectrum