MIMO Systems Spatial Multiplexing Diversity Capacity

Understanding MIMO wireless communication including spatial multiplexing for throughput, spatial diversity for reliability, channel capacity formulas, precoding techniques, and implementation in LTE and 5G.

MIMO Channel Model

The Channel Matrix

For a system with Nt transmit antennas and Nr receive antennas, the channel is described by an Nr × Nt matrix H:

Each element hij represents the complex channel gain from transmit antenna j to receive antenna i. The received signal vector:

y = Hx + n

Where x is the Nt×1 transmitted signal vector, n is the Nr×1 noise vector.

Channel Rank

The rank of H determines how many independent spatial streams can be supported. In a rich scattering environment, rank = min(Nt, Nr). In a pure line-of-sight channel (no scattering), rank = 1 regardless of antenna count. This is why MIMO works best in urban environments with abundant multipath — the very condition that was previously considered harmful.

MIMO Capacity

Single Antenna (SISO) Capacity

C_SISO = B × log₂(1 + SNR) bits/sec

Doubling SNR gives only 1 additional bit/sec/Hz (logarithmic — severely diminishing returns).

MIMO Capacity (Channel Known at Receiver)

C_MIMO = B × Σᵢ₌₁ʳ log₂(1 + λᵢ × P/(Nt × N₀)) bits/sec

Where λᵢ are eigenvalues of HHᴴ, r = rank(H).

In high-SNR regime with full rank channel: C_MIMO ≈ min(Nt, Nr) × B × log₂(SNR/Nt) bits/sec

This is a LINEAR increase with antenna count — revolutionary compared to SISO's logarithmic improvement with power.

Capacity Scaling Example

Spatial Multiplexing

Concept

Spatial multiplexing transmits independent data streams from different antennas simultaneously on the same frequency. The receiver separates these streams using knowledge of the channel matrix H.

Detection Algorithms

V-BLAST Architecture

V-BLAST (Vertical Bell Labs Layered Space-Time) was the first practical spatial multiplexing scheme:

1. Independent data streams sent from each antenna (no coding between streams)
2. Receiver detects the stream with highest SINR first
3. Detected stream is subtracted from received signal (interference cancellation)
4. Next strongest stream detected from residual signal
5. Process repeats until all streams recovered

V-BLAST achieved 40 bps/Hz spectral efficiency in lab demonstrations — far beyond any previous wireless system.

Spatial Diversity

Concept

Instead of sending different data on different antennas (multiplexing), diversity sends the SAME data across multiple paths. If one path fades, others likely remain strong — dramatically reducing the probability of deep fades.

Diversity Gain

The diversity order d determines how quickly error probability decreases with SNR:

• Without diversity: P_error ∝ 1/SNR (linear decrease)
• With diversity order d: P_error ∝ 1/SNR^d (much faster decrease)

For Nr receive antennas: Maximum diversity order = Nr For Nt×Nr MIMO: Maximum diversity order = Nt × Nr

Space-Time Codes

The Alamouti code is remarkable: it achieves full transmit diversity (order 2) with full rate and requires only linear decoding at the receiver. It is used in LTE's transmit diversity mode.

Diversity-Multiplexing Trade-off (DMT)

The Fundamental Trade-off

Zheng and Tse (2003) proved that diversity gain and multiplexing gain cannot both be maximized simultaneously. The achievable trade-off is:

d(r) = (Nt - r)(Nr - r) for integer multiplexing gains r

Where r = multiplexing gain (number of independent streams), d = diversity gain.

For a 2×2 system:

• r = 0: d = 4 (maximum diversity, zero multiplexing — like Alamouti)
• r = 1: d = 1 (one spatial stream, minimal diversity)
• r = 2: d = 0 (maximum multiplexing, zero diversity — like V-BLAST)

Practical Implication

Modern systems adaptively switch between multiplexing and diversity based on channel conditions:

• Good channel (high SNR, high rank): Use spatial multiplexing for maximum throughput
• Poor channel (low SNR, fading): Use diversity for reliability
• LTE dynamically selects between MIMO modes based on reported CQI and rank indicator

MIMO in Standards

LTE MIMO

5G NR MIMO

5G NR supports up to 256 antenna elements and 8 MIMO layers per user (Release 15), with Type I and Type II codebooks for CSI reporting. Multi-user MIMO serves up to 12 users simultaneously on the same time-frequency resource.

Key Takeaways

1. MIMO provides linear capacity scaling with min(Nt, Nr) — a revolutionary improvement over SISO's logarithmic scaling with power
2. Spatial multiplexing sends independent streams from different antennas simultaneously, multiplying throughput without additional bandwidth
3. Spatial diversity sends redundant information across multiple paths, dramatically reducing outage probability in fading channels
4. The diversity-multiplexing trade-off (DMT) proves these two gains cannot be simultaneously maximized — systems must choose based on channel conditions
5. MIMO works best in rich scattering environments where the channel matrix has full rank (all paths are independent)
6. The Alamouti space-time block code achieves full diversity from 2 Tx antennas with linear decoding complexity — used in LTE transmit diversity
7. Modern systems (LTE, 5G) adaptively switch between multiplexing and diversity modes based on real-time channel quality measurements