Wireless communication systems must allow many users to share limited radio spectrum while remaining reliable in the presence of noise, fading, interference, intentional jamming, and imperfect synchronization. Spread-spectrum techniques address these problems by deliberately expanding a signal over a bandwidth substantially larger than the minimum required for its information rate. Multiple-access techniques exploit time, frequency, code, space, or combinations of these dimensions to permit simultaneous transmissions. Interference management then coordinates resources and receiver processing so that undesired signals do not prevent reliable detection.

These subjects are closely related. Code-division multiple access (CDMA), for example, uses spreading codes both to separate users and to provide processing gain. Modern cellular systems generally use orthogonal frequency-division multiple access (OFDMA), spatial beamforming, scheduling, interference coordination, and sometimes non-orthogonal transmission. The underlying principles remain essential for the design of 4G, 5G, satellite, military, sensor, and industrial communication systems.

Learning Objectives

1. Spread-Spectrum Fundamentals

A conventional digitally modulated signal occupies a bandwidth related primarily to its symbol rate and pulse-shaping filter. A spread-spectrum transmitter introduces a spreading waveform or hopping pattern that increases the occupied bandwidth. At the receiver, a synchronized despreading operation restores the desired signal while reducing the relative effect of some interference.

Spread spectrum is not identical to encryption. A spreading code may obscure a waveform from an uninformed receiver, but confidentiality requires cryptographic protection.

1.1 Processing Gain

Let R_c be the chip rate and R_b the information bit rate. For direct-sequence spread spectrum, the processing gain is approximately

where B_SS is the spread bandwidth and B_N is the bandwidth of an equivalent narrowband signal. In decibels,

For a binary phase-shift keying signal with one bit represented by N_c chips, R_c = N_cR_b and therefore G_p = N_c. The approximation assumes comparable pulse-shaping factors and that the spread bandwidth is dominated by the chip waveform rather than by excess filtering.

Processing gain is not an unconditional increase in received signal power. It is a receiver signal-processing advantage: after despreading, desired energy is coherently accumulated, whereas sufficiently uncorrelated interference is distributed over the decision bandwidth.

1.2 Direct-Sequence Spread Spectrum

In direct-sequence spread spectrum (DSSS), each information symbol is multiplied by a high-rate pseudonoise (PN) sequence. Represent a bipolar data symbol by d ∈ {+1, −1} and the chip sequence by c_k ∈ {+1, −1}. During one symbol interval, the transmitted chips are

where E_b is the energy per information bit. The factor ensures that the total energy across N_c chips remains E_b, under the stated rectangular-chip model.

A synchronized correlator multiplies received chips by c_k and sums them. Since c_k² = 1, the desired terms add coherently. An interfering narrowband signal generally does not have the same chip pattern and is reduced, although the exact residual depends on correlation, spectral shape, timing, and receiver filtering.

The spreading sequence may be periodic or very long. Important sequence properties include balance, autocorrelation, cross-correlation, linear complexity, and ease of synchronization. Gold sequences, m-sequences, Kasami sequences, and orthogonal codes have been used in different systems, but no sequence family simultaneously provides perfect correlation properties for all delays and all users.

1.3 Frequency-Hopping Spread Spectrum

In frequency-hopping spread spectrum (FHSS), the carrier frequency changes according to a pseudorandom hopping pattern. A hop may last for one or several symbols. In slow hopping, several symbols occur per hop; in fast hopping, several hops occur during one symbol.

FHSS provides frequency diversity and can avoid a narrowband interferer if the hop set is sufficiently broad. Its principal design parameters include the number of hop channels, hop duration, synthesizer settling time, synchronization accuracy, and hop-pattern security. A receiver must know the sequence and timing or perform acquisition over possible patterns.

DSSS and FHSS differ operationally. DSSS expands the signal continuously in frequency, whereas FHSS relocates a narrower waveform among frequency channels. Hybrid systems combine direct sequence, frequency hopping, time hopping, coding, or interleaving.

1.4 Time-Hopping and Ultra-Wideband Signaling

Time-hopping assigns pulses or symbols to selected positions within a frame. It is useful in impulse-radio ultra-wideband systems and can support multiple users through distinct time patterns. The main difficulties are timing acquisition, multipath resolution, regulatory spectral masks, and sensitivity to oscillator and sampling errors.

1.5 Despreading and Narrowband Interference

Suppose a desired DSSS signal is received with energy E_b and a narrowband interferer has power spectral density approximately constant over the spread bandwidth. After despreading, only the fraction corresponding to the information bandwidth contributes significantly to the decision variable. An idealized relation is

where S/J denotes desired-signal-to-jammer ratio. This relation is an engineering approximation. It assumes synchronization, a linear receiver, an approximately uncorrelated interferer, and no severe nonlinearities or front-end blocking. A matched narrowband interferer using the same spreading code is not suppressed in this way.

2. Multiple-Access Principles

Multiple access concerns how independent transmitters share a common medium. It must be distinguished from multiplexing: multiplexing often combines streams from one transmitter, whereas multiple access coordinates several transmitters using a shared resource.

2.1 FDMA and TDMA

In frequency-division multiple access (FDMA), each user receives a separate frequency channel, usually with guard bands. FDMA is conceptually simple and supports continuous transmission, but guard bands reduce spectral efficiency and frequency-selective propagation may affect users unequally.

In time-division multiple access (TDMA), users transmit in assigned time slots. A frame contains multiple slots, often with guard times to accommodate propagation delay and timing uncertainty. TDMA requires accurate timing and introduces delay when a user waits for its slot, but it enables flexible burst allocation and sleep intervals.

2.2 CDMA

In code-division multiple access, users occupy overlapping time and frequency resources but employ distinguishable spreading codes. For a synchronous system with normalized user signatures s_i, the received vector can be written as

where P_i is received power, b_i is the transmitted symbol, s_i is the spreading signature, and n is additive noise. The cross-correlation ρ_ij = s_iᵀs_j determines how much user j interferes with user i after correlation detection.

where N₀B is the in-band noise power under a white-noise model. The expression assumes normalized signatures, synchronous symbol timing, and a simplified flat channel. In asynchronous multipath channels, effective cross-correlations include timing offsets, channel responses, and multipath components.

2.3 OFDM and OFDMA

Orthogonal frequency-division multiplexing (OFDM) divides a high-rate stream among many closely spaced orthogonal subcarriers. In OFDMA, different subcarriers or resource blocks are assigned to different users. Orthogonality is maintained when subcarrier spacing, sampling, frequency offset, and timing satisfy system requirements.

A cyclic prefix converts linear convolution with a sufficiently short channel into approximately circular convolution, enabling simple frequency-domain equalization. The prefix consumes overhead. If the useful symbol duration is T_u and the cyclic-prefix duration is T_cp, the efficiency due only to the prefix is

This assumes that other pilots, guard bands, coding, and control overhead are excluded from the expression.

OFDMA is highly compatible with frequency-selective scheduling: users can be assigned resource blocks where their channel gains are favorable. Its limitations include high peak-to-average power ratio, sensitivity to carrier-frequency offset and phase noise, and inter-carrier interference when orthogonality is lost.

2.4 Spatial-Division and Non-Orthogonal Access

Spatial-division multiple access (SDMA) separates users through antenna arrays, beamforming, and spatial signatures. With M antennas, a base station can exploit channel vectors to serve several users, although the useful number of streams depends on channel rank, correlation, signal-to-noise ratio, and hardware constraints.

Non-orthogonal multiple access (NOMA) deliberately allows users to share a resource and separates them by power allocation, code structure, or advanced receivers. In power-domain NOMA, successive interference cancellation (SIC) decodes a relatively strong or weak user first, subtracts its reconstructed signal, and decodes another. Error propagation, channel ordering, fairness, and signaling overhead limit practical gains. NOMA should therefore be evaluated against a well-designed orthogonal scheduler rather than assumed to be universally superior.

2.5 Comparison of Multiple-Access Methods

Method	Separation dimension	Main strength	Principal limitation
FDMA	Frequency	Simple continuous access	Guard bands and inflexible allocation
TDMA	Time	Flexible burst scheduling	Timing and slot delay
CDMA	Code/signature	Soft capacity and frequency reuse	Near–far problem and interference accumulation
OFDMA	Orthogonal subcarriers	Fine frequency-selective scheduling	Synchronization and peak power issues
SDMA	Spatial channel	Reuses time-frequency resources	Channel estimation and antenna complexity
NOMA	Power/code and receiver processing	Potential overload and flexible multiplexing	SIC complexity and error propagation

3. Interference: Models and Metrics

Noise is generally modeled as an unwanted random process generated by physical components. Interference is energy from other transmitters, reflections, adjacent channels, or intentional jammers. Unlike thermal noise, interference may be structured, correlated, directional, or adaptive.

where S, I, and N are received desired-signal, interference, and noise powers in watts. In decibels, powers must first be converted consistently; in general, dB values cannot be added as ordinary linear quantities. For a target SINR γ_req and transmit power P_t, a simplified link constraint is

where G_t and G_r are antenna gains and L is the dimensionless path-loss factor. This model omits fading margins and implementation loss unless they are incorporated into γ_req or L.

3.1 The Near–Far Problem

In CDMA, a nearby transmitter may arrive at a receiver with much greater power than a distant desired user. Even if their codes have low cross-correlation, the strong user’s residual correlation can overwhelm the weak signal. This is the near–far problem.

Power control attempts to make received powers compatible with target quality. Open-loop control estimates path loss; closed-loop control uses feedback from the receiver. Perfect equalization is neither always possible nor always desirable: users have different quality requirements, channel uncertainty, and power constraints.

3.2 Interference Temperature and Capacity

A system may treat aggregate interference as an effective noise level, but this approximation fails when interference is strong, structured, or decodable. Capacity and outage depend on the distribution of SINR, not only its average. Interference-aware design therefore considers percentiles, spatial correlation, temporal variation, and the receiver’s ability to exploit structure.

4. Interference-Management Techniques

4.1 Resource Allocation and Coordination

Interference can be reduced by assigning different frequencies, time slots, codes, beams, or transmit powers. In cellular networks, frequency reuse, fractional frequency reuse, coordinated scheduling, and coordinated beamforming manage inter-cell interference. A scheduler may maximize throughput, proportional fairness, weighted fairness, or reliability subject to power and quality constraints.

where R_i is user rate, U_i is a utility function, w_i is a priority weight, p_i is allocated power, P_max is a power budget, and γ_i,min is a quality constraint. The problem is often non-convex because SINR variables are coupled through interference. Practical systems use decomposition, iterative optimization, heuristics, or machine-learning-assisted policies.

4.2 Filtering, Guard Intervals, and Waveform Design

Transmit and receive filters limit adjacent-channel leakage. Raised-cosine or root-raised-cosine filtering controls bandwidth and intersymbol interference. Guard bands and guard times provide robustness at the expense of spectral efficiency. OFDM systems use windowing, subcarrier mapping, and synchronization algorithms to limit leakage.

4.3 Receive Processing

A matched filter maximizes output SNR in additive white Gaussian noise for a known waveform. In interference-limited environments, a linear minimum mean-square error (MMSE) receiver can provide better performance by balancing noise enhancement and interference suppression. In matrix form, for received vector r = Hb + n with covariance R_nn, an MMSE linear estimator has the form

where H is the channel matrix, Hᴴ is its conjugate transpose, σ² is a noise-scale parameter, and R_b is the transmitted-symbol covariance. The exact expression depends on the adopted signal and noise normalization. Zero-forcing detection is a limiting interference-nulling approach but can strongly amplify noise when H is ill-conditioned.

Multiuser detection jointly estimates users rather than treating all others as noise. Maximum-likelihood detection can approach optimum performance but has exponential complexity in the number of users or streams. Sphere decoding, message passing, successive interference cancellation, and iterative turbo receivers provide practical compromises.

4.4 Successive Interference Cancellation

SIC estimates one signal, reconstructs it using the estimated channel and symbols, subtracts it from the received waveform, and proceeds to another signal. Ordering may be based on received power, channel quality, reliability, or an optimization criterion. Cancellation errors propagate, so robust systems use soft decisions, reliability tests, coding, and sometimes parallel cancellation.

4.5 Adaptive and Cooperative Methods

Adaptive beamforming places spatial nulls toward interferers while preserving gain toward the desired user. Cooperative relaying and coordinated multipoint transmission allow multiple nodes to share channel information or jointly transmit. These techniques improve cell-edge performance but require backhaul capacity, synchronization, channel-state information, and careful privacy and security controls.

5. Worked Example: Processing Gain and SINR

A DSSS link transmits at R_b = 100 kbit/s using a chip rate R_c = 10 Mchip/s. Assume the spread bandwidth is approximately the chip rate and that an uncorrelated narrowband interferer is present. The processing gain is

If the input desired-signal-to-interference ratio is −15 dB, the idealized output ratio is approximately

This result does not guarantee a 5 dB link margin. Coding loss, synchronization error, multipath, code cross-correlation, receiver implementation loss, and interference correlation may substantially reduce the improvement. The example illustrates processing gain, not a universal performance guarantee.

6. Applied Case: Interference Management in a Dense Industrial Wireless Network

Background

Consider an illustrative factory containing autonomous vehicles, machine sensors, safety controllers, and video terminals. Metal structures produce multipath, vehicles create shadowing, and many devices transmit in the same indoor area. Safety traffic requires low latency and high reliability, whereas video traffic is delay tolerant.

Problem

A single static channel allocation causes collisions during machine-control bursts. Increasing transmit power improves some links but raises interference at neighboring machines. A CDMA-only solution suffers from near–far effects, while an uncoordinated OFDMA solution experiences adjacent-resource leakage and synchronization failures from mobile terminals.

Analysis

Measurements are used to estimate channel gain, interference maps, traffic periodicity, and latency distributions. Safety packets are assigned strict reliability and latency constraints. The network controller identifies highly correlated interference zones and distinguishes persistent interferers from short impulsive events.

Approach

A fallback contention channel is reserved for emergency access. The controller evaluates both mean throughput and 99th-percentile latency, because average throughput alone can conceal dangerous reliability failures.

Outcome

The hybrid design reduces avoidable co-channel interference and separates traffic according to service requirements. Its cost is greater channel-estimation, scheduling, and synchronization complexity. The principal engineering lesson is that interference management is a systems problem: waveform choice, resource allocation, antenna design, coding, power control, and traffic policy must be considered together.

7. Design, Regulatory, and Security Considerations

Spread-spectrum systems must comply with national spectrum regulations concerning occupied bandwidth, out-of-band emissions, power spectral density, duty cycle, and coexistence. Regulatory classification differs across frequency bands and jurisdictions; engineers must consult current applicable rules rather than rely on generic descriptions.

Security considerations include protection of spreading and hopping patterns, authentication of control messages, resilience to replay and spoofing, and detection of intentional jamming. A secret hopping pattern is not a substitute for cryptographic authentication. Machine-learning-based interference classifiers can identify spectral anomalies, but they require representative training data, protection against adversarial examples, explainability for safety decisions, and fallback behavior when confidence is low.

Implementation impairments include oscillator phase noise, carrier-frequency offset, sampling-clock offset, nonlinear power amplifiers, quantization, finite-resolution channel estimates, timing drift, and imperfect code synchronization. Peak-to-average power ratio is especially significant in OFDM; amplifier back-off improves linearity but reduces power efficiency.

8. Contemporary Directions and Critical Perspective

Current research combines massive multiple-input multiple-output (MIMO), intelligent scheduling, reconfigurable surfaces, integrated sensing and communication, cell-free architectures, and AI-assisted radio resource management. These approaches promise better spatial reuse and adaptability, but their benefits depend on channel-state acquisition, control overhead, hardware cost, and realistic mobility models.

A recurring research mistake is to compare a sophisticated method with an unrealistic baseline. Meaningful evaluation should report bandwidth, transmit power, coding, channel model, synchronization assumptions, computational complexity, pilot overhead, fairness, outage probability, and robustness to model mismatch. Simulation should be complemented by analytical bounds, hardware experiments, or field measurements where possible.

Chapter Summary

Spread spectrum deliberately enlarges signal bandwidth to obtain processing gain, frequency diversity, interference tolerance, or user separation. DSSS multiplies data by high-rate codes, while FHSS changes carrier frequency according to a hopping pattern. Processing gain is approximately the ratio of spread bandwidth to information bandwidth, but its practical benefit depends on synchronization, correlation, channel conditions, and receiver implementation.

FDMA, TDMA, CDMA, OFDMA, SDMA, and NOMA separate users in frequency, time, code, orthogonal subcarriers, space, or receiver-assisted shared resources. SINR is the central quality metric, and interference may be co-channel, adjacent-channel, intersymbol, inter-carrier, inter-cell, self-generated, or intentional. Power control addresses the CDMA near–far problem; filtering, scheduling, beamforming, MMSE detection, multiuser detection, SIC, and coordinated transmission provide complementary remedies.

Practical design requires joint consideration of spectral efficiency, reliability, latency, energy, complexity, synchronization, regulation, and security. No access method is universally optimal; the correct choice depends on traffic, channel structure, mobility, antenna resources, and interference geometry.

Key Terms

Spread spectrum: Transmission technique that intentionally occupies more bandwidth than the minimum information bandwidth.

Processing gain: Ratio of spread bandwidth to unspread information bandwidth, often expressed in decibels.

DSSS: Direct-sequence spread spectrum using a high-rate code to multiply the data waveform.

FHSS: Frequency-hopping spread spectrum using a pseudorandom sequence of carrier frequencies.

PN sequence: Deterministic sequence with noise-like statistical and correlation properties.

Multiple access: Sharing of communication resources among independent transmitters.

Near–far problem: Excessive interference caused when a strong transmitter masks a weak desired user.

Power control: Adjustment of transmit power to satisfy quality and interference objectives.

Multiuser detection: Joint or structured detection of multiple simultaneous signals.

SIC: Successive interference cancellation by estimating and subtracting signals sequentially.

Inter-cell interference: Interference generated by transmissions in neighboring cells.

Processing gain: Receiver improvement obtained through correlation or despreading under suitable interference assumptions.

Review Questions

Short Answer Questions

Long Answer Questions

Analytical / Application Questions

Multiple Choice Questions with Answers

Question 1

A DSSS system has a chip rate of 8 Mchip/s and an information bit rate of 80 kbit/s. What is its approximate processing gain?

Question 2

Question 3

In an idealized spread-spectrum receiver, processing gain most directly improves the desired-signal-to-jammer ratio when the jammer is:

Question 4

Question 5

Question 6

Question 7

C. To reduce the effect of intersymbol interference and enable circular convolution under suitable conditions

Correct Answer: C. To reduce the effect of intersymbol interference and enable circular convolution under suitable conditions

Question 8

Correct Answer: B. Estimating, reconstructing, and subtracting signals sequentially

Chapter 10 Spread Spectrum, Multiple Access and Interference Management

Spread Spectrum, Multiple Access and Interference Management

Introduction and Context