ELEC5508 Chap.6 Multiuser Diversity and Scheduling
Multiuser Diversity and Scheduling
This chapter turns independent user fading into a scheduling resource. A max-rate scheduler follows the strongest instantaneous channel, while a normalised metric can reward a user whose current state is strong relative to its own average. Opportunistic beamforming can create useful variation in a slow environment, but feedback delay, control overhead, queue state and fairness limit the gain.
Selection benefit is statistical and depends on a population of eligible users with sufficiently distinct channel traces. It is not ordinary transmit diversity because different users receive different data. It is also not free capacity: channel reports consume control resources and can age before the scheduler acts.
A useful evaluation therefore reports sum throughput together with tail rate, delay or a fairness measure rather than celebrating the largest instantaneous peak alone.
What this chapter covers
- 01
Independent user peaks and selection gain
- 02
Max-rate scheduling
- 03
Normalised opportunity and proportional fairness
- 04
Feedback timing and channel ageing
- 05
Opportunistic beamforming
- 06
Opportunistic nulling in cellular interference
Worked example · free
Compare absolute and normalised selection
- +1Max-rate selects user 1 from the largest current rate of 12.
- +1Normalised metrics are [1.2, 2.0, 2.5].
- +1The normalised scheduler selects user 3.
- +1User 3 has the lowest absolute rate but the strongest relative opportunity.
Key terms
- Multiuser diversity
- System gain from selecting among users with independently varying channel quality.
- Max-rate scheduler
- A scheduler choosing the eligible user with the largest instantaneous supported rate.
- Proportional fairness
- A balance that rewards current quality relative to a user's longer-term service level.
- Channel feedback
- Receiver information sent to the scheduler about current link quality or direction.
- Opportunistic beamforming
- Deliberate beam variation that creates channel peaks for multiuser selection.
Multiuser Diversity and Scheduling FAQ
Why does multiuser diversity improve with more users?
With independent channel variation, more eligible users increase the chance that at least one has a strong instantaneous state. The gain depends on timely feedback and distinct fading; identical or stale reports erase the selection opportunity.
Why can max-rate scheduling be unfair?
Users with favourable average geometry can dominate every comparison, leaving edge users backlogged. A normalised, deficit-based or delay-aware metric includes service history or urgency so weaker users are scheduled when their own channels are relatively favourable.
When is opportunistic beamforming helpful?
It is useful when many users share a slowly varying environment that otherwise offers little temporal selection. Random beam changes create peaks for different users. A single delay-critical user cannot wait for those peaks and may be harmed by the added variation.
How should I evaluate an opportunistic scheduler?
Compute the stated selection metric only for users with queued data, then record the chosen user and update any historical average used by the rule. Compare max-rate and normalised selection on the same channel traces so throughput and fairness effects are attributable to the metric. Run enough independent fading realisations and user populations to avoid mistaking one favourable trace for a general gain.
Report sum throughput together with a lower-percentile user rate, fairness index or service gap. Include feedback quantisation, reporting delay and mobility because stale channel information can reverse the intended choice. Finally, state whether delay constraints permit waiting for a peak; a scheduler that raises average rate can still be unsuitable for urgent traffic.
Exam move
Sketch several independent channel traces and the selected upper envelope. Then compare max-rate with normalised metrics, always naming fairness, feedback age and queue state as design boundaries. Begin with a two-user example and identify which user is strongest in every slot. Add users and observe why the upper envelope rises when fades are independent.
Repeat with perfectly correlated traces to show that user count alone creates no selection gain. Work scheduler calculations with current rates and historical averages. Compute both absolute and normalised metrics, then update the selected user's average so the long-run feedback mechanism becomes visible. Add queue eligibility and deadline weights only after the channel metric is correct.
For opportunistic beamforming, explain why deliberate beam variation helps a multiuser slow-fading system and why it can harm a single delay-sensitive link. For opportunistic nulling, inspect both desired power and interference because the favourable event sits in the SINR numerator and denominator together. When reviewing a graph, check how many users, channel realisations and feedback delays were simulated.
Compare against round-robin service under the same traffic and propagation model. Report average throughput, a lower-percentile user rate and a fairness index or service gap. Conclude with the control cost: feedback bits, update interval and channel ageing. This keeps a beautiful peak-rate curve from hiding starvation or stale decisions.
Working through Multiuser Diversity and Scheduling in ELEC5508? Sia is AskSia’s AI Electrical Engineering tutor — ask any ELEC5508 Multiuser Diversity and Scheduling question and get a clear, step-by-step explanation grounded in how ELEC5508 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.