ECE4024 Chap.8 Diversity Combining
Diversity Combining
Diversity uses independently faded observations to reduce the chance that every copy is weak at once. This chapter compares selection, equal-gain and maximal-ratio combining, keeps linear and logarithmic SNR arithmetic separate, and distinguishes array gain from diversity order. It also connects receive combining to transmit diversity and explains why correlation and estimation error limit ideal antenna-count claims.
What this chapter covers
- 01
Space, time and frequency diversity
- 02
Independent fading and outage
- 03
Selection combining
- 04
Equal-gain coherent combining
- 05
Maximal-ratio combining
- 06
Linear SNR addition and decibel conversion
- 07
Array gain versus diversity order
- 08
Correlation, estimation error and transmit diversity
Worked example · free
Compare selection and maximal-ratio combining
- +1Selection combining takes the largest branch: gamma-SC = 6.5.
- +1Convert selection to decibels: 10log10(6.5) is about 8.13 dB.
- +2MRC adds linear branch SNRs: 1.5 + 4 + 6.5 = 12.
- +1Convert MRC to decibels: 10log10(12) is about 10.79 dB.
- +1The realization gain over selection is about 2.66 dB.
- +1State the assumptions of coherent weights, sufficiently independent branches and the standard noise model.
Key terms
- Selection combining
- A receiver rule that detects from the branch with the largest measured instantaneous SNR.
- Equal-gain combining
- A coherent rule that aligns branch phases and adds them with equal magnitude weights.
- Maximal-ratio combining
- A coherent rule that weights branches according to complex channel quality and maximises output SNR under the standard model.
- Array gain
- An improvement in effective SNR from coherently using several observations or antennas.
- Diversity order
- The high-SNR slope of outage or error probability associated with independent fading dimensions.
- Alamouti coding
- An orthogonal two-transmit-antenna space-time structure that provides diversity under its channel assumptions and a corresponding receiver combiner.
Diversity Combining FAQ
Why is MRC stronger than selection combining?
Selection discards all branches except the best. MRC co-phases and weights every useful observation, so under the standard independent-noise model its output SNR is the sum of branch SNR values.
Why must branch SNR values be added in linear form?
SNR is a power ratio and independent power contributions add linearly. Decibels are logarithmic representations, so adding decibel numbers would multiply underlying ratios rather than sum the received evidence.
Is diversity order the same as antenna count?
Only under suitable independence, channel knowledge and combining assumptions. Strongly correlated antennas observe similar fades and provide fewer independent reliability dimensions, even if they still contribute some coherent array gain.
How does transmit diversity differ from beamforming?
Transmit diversity arranges redundant symbol evidence across antennas or time to improve reliability. Beamforming uses channel-aware weights to align energy in a chosen direction, primarily producing coherent array gain for a stream.
Why is raising transmit power not equivalent to diversity?
A uniform power increase shifts SNR but does not create an alternative observation when a path is deeply faded. Diversity changes the probability of simultaneous failure by providing separate fading dimensions.
Exam move
Practise converting branch SNRs from decibels to linear units, applying selection and MRC rules, and converting the combined result back. State the independence and channel-knowledge assumptions beside every ideal diversity-order claim. Draw receive branches with phase alignment and weights so EGC and MRC are visually distinct.
Compare transmit diversity, beamforming and spatial multiplexing by objective: reliability, coherent gain and parallel rate. Use outage language only when a threshold and probability model are actually given. Create one receiver diagram with three branch paths and reuse it to test each combining rule. Selection combining keeps the branch with the strongest instantaneous quality.
Equal-gain combining corrects phases and uses equal magnitudes. Maximal-ratio combining corrects phases and weights more reliable observations more heavily under the standard noise assumptions. Label where channel estimates and noise information enter so the apparent algebra corresponds to hardware knowledge.
In numerical drills, convert all branch SNR values to linear form before selecting or adding; adding dB values is not MRC. Check that ideal MRC output SNR is at least as large as the best individual branch and then convert the final result back to decibels. For outage work, write the threshold event first and specify the fading distribution and independence assumptions before multiplying probabilities.
Distinguish diversity order, which describes a high-SNR reliability slope, from array gain, which shifts performance. Correlated antennas or common blockage can reduce the independence benefit even when the branch count is unchanged. Rehearse transmit diversity separately: identify the coding across antennas and time, the receiver observations and the channel estimates required for combining.
Then contrast it with beamforming, which uses coherent transmit weights, and spatial multiplexing, which sends parallel information. End every comparison by naming the objective, channel knowledge, extra resources and principal failure mode. That structure prevents antenna count alone from deciding whether a system provides reliability or rate.
Working through Diversity Combining in ECE4024? Sia is AskSia’s AI Electrical Engineering tutor — ask any ECE4024 Diversity Combining question and get a clear, step-by-step explanation grounded in how ECE4024 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
