Channel access method
A channel access method, also known as a multiple access method, is a protocol or technique in telecommunications and computer networks that enables multiple devices or users to share a common communication channel or transmission medium efficiently, minimizing interference and collisions while optimizing resource utilization.[1] These methods are foundational to multiplexing, allowing several data streams or signals to coexist on the same physical medium, and are essential for systems ranging from wired local area networks (LANs) to wireless cellular networks.[2] Channel access methods are broadly categorized into orthogonal multiple access (OMA) and non-orthogonal multiple access (NOMA), with additional contention-based approaches for dynamic environments. OMA techniques assign distinct resources to users, such as frequency bands in frequency-division multiple access (FDMA), time slots in time-division multiple access (TDMA), or spreading codes in code-division multiple access (CDMA), which have powered generations of cellular systems from 1G to 4G.[3] NOMA, prominent in 5G and beyond, permits overlapping resource use through power differentiation or advanced coding, enhancing spectral efficiency and supporting massive device connectivity in scenarios like the Internet of Things (IoT).[4] Contention-based methods, such as carrier-sense multiple access with collision detection (CSMA/CD) in Ethernet or collision avoidance (CSMA/CA) in Wi-Fi, rely on devices sensing the medium before transmitting and employing backoff mechanisms to resolve conflicts in shared, uncoordinated settings.[2] The evolution of channel access methods reflects advancements in network demands, from early satellite and telephone systems using fixed assignments to modern hybrid schemes integrating centralized control (e.g., via a hybrid coordinator in IEEE 802.11e) with distributed access for quality-of-service (QoS) guarantees.[5] These techniques ensure fair bandwidth allocation, high throughput, and low latency, with ongoing research focusing on AI-driven adaptations for unknown or dynamic environments like vehicular networks.[6]Introduction
Definition and Core Principles
A channel access method, also referred to as a multiple access method, is a protocol or technique that enables multiple users or devices to share a single communication channel or transmission medium without significant interference. This sharing is facilitated by multiplexing principles, which combine multiple data streams or signals into a single transmission over the shared medium, allowing efficient resource utilization.[7] At its core, multiplexing divides the available channel resources along various dimensions, such as time, frequency, code, space, or power, to allocate portions to different users. Orthogonal multiple access approaches assign these resources in a non-overlapping manner to prevent interference, exemplified by techniques like frequency-division multiple access (FDMA) using distinct frequency bands or time-division multiple access (TDMA) employing time slots. In contrast, non-orthogonal methods permit controlled overlap and interference among signals, often decoded through advanced receiver processing, to achieve higher density of users. These principles involve inherent trade-offs: orthogonal methods typically offer lower implementation complexity and greater robustness to interference but may sacrifice bandwidth efficiency, while non-orthogonal schemes enhance spectral utilization at the expense of increased latency and processing demands.[8][9][10] The theoretical foundation for these methods stems from Shannon's capacity theorem, which defines the maximum reliable data rate over a channel as C = B \log_2 (1 + \mathrm{SNR}), where B is the bandwidth and \mathrm{SNR} is the signal-to-noise ratio. In multiple access contexts, this extends to a capacity region representing the set of achievable rate tuples for multiple users sharing the channel, bounded by mutual information constraints without exceeding the single-user limit. Resource dimensions include time slots for sequential access, frequency bands for parallel subchannels, orthogonal codes for signal distinction, spatial beams for directional separation, and power levels for layered allocation. These methods are crucial for optimizing spectrum efficiency in wireless networks, wired systems, and broadcast environments, supporting simultaneous transmissions from numerous devices.[11][12][13]Historical Development
The origins of channel access methods trace back to the 19th century, when multiplexing techniques emerged in wired telephony to enable multiple signals over shared lines. Early frequency-division multiplexing (FDM) concepts were proposed by Alexander Graham Bell around 1870 for harmonic telegraphy, allowing simultaneous transmission of tones at different frequencies on a single wire. By the early 1900s, these principles extended to radio communications, where frequency separation was used to avoid interference in pioneering wireless broadcasts and telephony experiments. Post-World War II advancements in radar technology further propelled multiple access innovations, as pulse techniques and signal processing from military applications were adapted for civilian broadcasting and early mobile radio systems, enhancing capacity in shared spectrum environments.[14][15][16] The first generation (1G) of cellular systems in the 1980s introduced frequency-division multiple access (FDMA) for analog voice services, marking the shift to mobile telephony. The Advanced Mobile Phone System (AMPS), launched commercially in the United States in 1983, divided spectrum into 30 kHz channels assigned exclusively to users, supporting basic voice calls but limited by spectrum inefficiency. Key drivers included growing demand for mobile connectivity amid spectrum scarcity, with regulatory actions like the U.S. Federal Communications Commission's (FCC) initiation of spectrum auctions in 1994 facilitating broader allocation for wireless services.[17][18][19] The 2G era in the 1990s transitioned to digital systems, emphasizing time-division multiple access (TDMA) for improved efficiency and security. The Global System for Mobile Communications (GSM), standardized in 1990 and first deployed in Finland in 1991, used TDMA to multiplex eight voice channels per 200 kHz carrier, becoming the dominant global standard. Concurrently, code-division multiple access (CDMA) emerged with the IS-95 standard, finalized by the Telecommunications Industry Association in 1995, enabling higher capacity through spread-spectrum techniques for simultaneous user access. These developments addressed escalating data rate needs and mobility requirements, with 3G systems in the 2000s building on CDMA via Universal Mobile Telecommunications System (UMTS), released by 3GPP in 1999 and commercially launched around 2001, incorporating wideband spread spectrum for multimedia support up to 2 Mbps.[20][21][22] The 4G era in the 2010s introduced orthogonal frequency-division multiple access (OFDMA) and single-carrier FDMA (SC-FDMA) through the Long-Term Evolution (LTE) standard, frozen by 3GPP in December 2008, to enable high-speed broadband data up to 100 Mbps downlink. LTE's adoption surged by 2010, driven by internet proliferation and spectrum auctions that expanded available bands. Entering the 5G era from 2019, enhanced hybrids integrated OFDMA with massive multiple-input multiple-output (MIMO) for spatial-division multiple access (SDMA), alongside non-orthogonal multiple access (NOMA) trials, mmWave bands for higher throughput, and low-latency protocols under 3GPP Release 15, supporting up to 20 Gbps and ultra-reliable communications.[23] Looking toward 6G beyond 2025, emerging paradigms focus on AI-optimized multiple access, rate-splitting multiple access (RSMA), and sensing-integrated techniques to handle terahertz frequencies and integrated sensing-communications. As of 2025, the European Telecommunications Standards Institute (ETSI) launched an Industry Specification Group on Multiple Access Techniques in January, exploring orthogonal, spatial, non-orthogonal, and rate-splitting methods aligned with 3GPP's 6G studies, which began formal workshops in March 2025 to address spectrum efficiency for holographic and AI-driven services. These evolutions continue to be propelled by persistent spectrum constraints, exponential data growth, and demands for seamless mobility.[24][25]Orthogonal Multiple Access Techniques
Frequency-Division Multiple Access
Frequency-division multiple access (FDMA) is an orthogonal multiple access technique that allocates discrete, non-overlapping frequency channels from the available spectrum to different users, enabling simultaneous transmission without mutual interference. Each user is assigned a specific frequency band for the duration of their communication session, typically managed by a central controller during call setup. To prevent crosstalk between adjacent channels, bandpass filters are employed at transmitters and receivers to confine signals within their allocated bands, while small guard bands—unused frequency gaps—are inserted between channels to further mitigate adjacent channel interference arising from filter roll-off imperfections or nonlinear distortions.[26] Variants of FDMA include fixed-channel allocation, as implemented in first-generation (1G) cellular systems like the Advanced Mobile Phone System (AMPS), where each user is permanently assigned a narrow frequency channel of 30 kHz within a total spectrum of 25 MHz, supporting up to 832 duplex channels with 45 MHz separation between uplink and downlink to avoid self-interference. In contrast, dynamic FDMA, often referred to as demand-assigned multiple access (DAMA), allows adaptive reallocation of frequency bands based on real-time traffic demands, improving flexibility in systems with variable user loads, such as certain satellite communications where channels are reassigned on demand to optimize resource use.[27][28] FDMA offers advantages in simplicity of implementation and low latency, making it suitable for constant bit rate (CBR) services like voice telephony, as users maintain continuous access to their dedicated channel without needing time synchronization. However, it suffers from spectrum inefficiency due to the overhead of guard bands, which can consume 10-20% of the total bandwidth, and vulnerability to frequency-selective fading, where multipath propagation affects specific frequency bands more severely than others, potentially degrading signal quality in mobile environments.[26][29] Mathematically, in FDMA, the total available bandwidth B is divided into N channels, with each channel having a usable width of approximately \frac{B}{N} - G, where G represents the guard band width per channel boundary to ensure sufficient separation. For interference mitigation, the adjacent channel power ratio (ACPR) is designed to provide attenuation greater than 60 dB, meaning the power leaking into the neighboring band is at least 60 dB below the in-band power, achieved through high-performance RF filters with sharp roll-off characteristics.[30] Early implementations of FDMA relied on analog modulation, such as frequency modulation (FM) in AMPS for voice transmission over 30 kHz channels, providing reliable but bandwidth-intensive service. Digital extensions have been applied in areas like satellite television broadcasting, where FDMA allocates distinct frequency bands within a transponder's spectrum to multiple carriers, enabling simultaneous delivery of various channels while maintaining orthogonality through precise frequency planning.[27][31] In terms of performance, basic FDMA achieves spectral efficiency of typically 0.5-1 bits/s/Hz, limited by guard bands and modulation overhead; for example, in a 25 MHz band divided into 1000 channels (each 25 kHz wide, assuming minimal guard bands of ~1-2 kHz), the effective efficiency per channel might yield around 0.8 bits/s/Hz for simple digital modulation like binary phase-shift keying (BPSK) at 12.5 kbps, highlighting the trade-off between user capacity and interference protection.[26][32]Time-Division Multiple Access
Time-division multiple access (TDMA) is an orthogonal channel access method that enables multiple users to share a single frequency channel by dividing the available time into discrete slots within a repeating frame structure. In this approach, each user is assigned one or more specific time slots per frame, during which they transmit bursts of data using the full channel bandwidth, while remaining silent otherwise to avoid interference. The frame structure typically consists of a fixed duration T, subdivided into N slots of length T/N for N users, with guard periods inserted between slots to account for propagation delays and switching transients. Synchronization is critical in TDMA systems and is achieved through preambles—short known bit sequences at the start of each burst that allow receivers to align timing and carrier phase—or external references like GPS for global coordination in satellite or wide-area networks. Burst transmissions are formatted to fit precisely within the allocated slots, often including header, data payload, error-correction coding, and tail bits for clean transitions.[33][34][35] TDMA variants include fixed and dynamic allocations to accommodate different traffic patterns. Fixed TDMA assigns predetermined slots to users regardless of demand, as seen in the Global System for Mobile Communications (GSM), where each 200 kHz carrier frame lasts 4.615 ms and contains 8 equal slots of approximately 577 μs, supporting up to 8 users per carrier for voice or data services. Dynamic TDMA, in contrast, adjusts slot assignments frame-by-frame based on variable bit-rate traffic, allowing unused slots to be reallocated for efficiency in bursty applications like packet data. Implementations of TDMA appear in digital cordless and cellular systems, such as the Digital Enhanced Cordless Telecommunications (DECT) standard, finalized in 1992 by the European Telecommunications Standards Institute (ETSI), which uses 10 ms frames with 24 time slots (12 for downlink and 12 for uplink) per 1.152 MHz carrier for short-range voice and data in cordless phones. Similarly, the IS-136 standard, an evolution of Digital AMPS for North American cellular networks, employs TDMA with 6 slots per 30 kHz frame to triple capacity over analog systems, supporting digital voice at 8 kbps per user.[36][37][38][39] A key advantage of TDMA is that each user accesses the full bandwidth during their slot, enabling high peak rates and efficient support for bursty data traffic without constant transmission power, which reduces interference and battery drain compared to continuous schemes. However, it incurs high synchronization overhead due to precise timing requirements, potentially leading to slot wastage from guard times, and introduces latency as users wait for their assigned slots, limiting suitability for delay-sensitive real-time applications. Mathematically, for a frame of duration T and N users, the slot length is T/N, yielding a duty cycle of $1/N per user; the aggregate throughput R is given by R = \frac{B \times N}{T}, where B is the data bits per slot, though effective throughput accounts for overhead like preambles and guards. In GSM, frame efficiency exceeds 90% with optimized burst structures, as guard and training sequences occupy less than 10% of the slot, enabling reliable operation at 13 kbps full-rate voice per user across 8 slots.[40][36]Non-Orthogonal and Code-Based Techniques
Code-Division Multiple Access
Code-division multiple access (CDMA) is a channel access method that enables multiple users to share the same frequency band and time resources simultaneously by assigning each user a unique spreading code, allowing the receiver to distinguish signals through despreading. In direct-sequence CDMA (DS-CDMA), the primary mechanism involves spreading the data signal across a wider bandwidth using a pseudorandom noise (PN) sequence or orthogonal codes like Walsh codes, where the chip rate exceeds the bit rate, creating a spread-spectrum signal. At the transmitter, the data bits are multiplied by the spreading code to generate a high-rate chip sequence; the receiver then uses a matched filter correlated with the same code to despread the signal, collapsing it back to the original bit rate while suppressing interference from other users' codes. This process relies on the near-orthogonal properties of the codes to minimize cross-correlation, but the near-far problem—where stronger signals from nearby users overwhelm weaker ones from distant users—necessitates power control mechanisms, such as closed-loop adjustments, to equalize received powers and maintain fair access.[41][41][42] CDMA variants include synchronous CDMA, which employs orthogonal Walsh codes for downlink scenarios where timing alignment is feasible, ensuring zero cross-correlation among codes within the same cell, and asynchronous CDMA, which uses longer PN sequences for uplink communications to handle timing offsets between users. Walsh codes, derived from Hadamard matrices, satisfy the orthogonality condition \sum_{i=0}^{N-1} w_k(i) w_m(i) = N \delta_{km}, where w_k and w_m are code sequences of length N, \delta_{km} is the Kronecker delta, allowing perfect separation in synchronized environments. In contrast, PN sequences provide pseudo-orthogonality for asynchronous operation but introduce some multiple-access interference due to non-zero cross-correlations.[41][41][41] The advantages of CDMA include robust resistance to interference and jamming through the spread-spectrum processing gain, soft capacity that increases gradually with load rather than abruptly like in TDMA or FDMA, and enhanced multipath diversity via rake receivers that combine delayed signal replicas. However, it suffers from disadvantages such as the need for complex receivers to handle multiuser detection and mitigate self-interference from imperfect orthogonality, particularly in asynchronous modes. The spreading factor [SF](/page/SF), defined as the ratio of chip rate to bit rate ([SF](/page/SF) = R_c / R_b), quantifies the bandwidth expansion, with processing gain given by $10 \log_{10}([SF](/page/SF)) dB; for example, with [SF](/page/SF)=64, the gain is 18 dB, improving signal-to-noise ratio against narrowband interference. System capacity in terms of maximum users K approximates K \approx (W/R) \cdot (E_b/N_0)^{-1}, where W is the chip-rate bandwidth, R is the user bit rate, and E_b/N_0 is the required energy per bit to noise power spectral density ratio for acceptable error rates, highlighting CDMA's interference-limited nature.[41][41][41][43] Practical implementations of CDMA include the IS-95 standard (also known as cdmaOne), released in 1995 by the Telecommunications Industry Association (TIA), which uses DS-CDMA with a 1.25 MHz bandwidth, Walsh codes for channelization, and PN sequences for user separation, achieving voice capacities up to 40-50 users per cell under typical conditions. Wideband CDMA (WCDMA), adopted in the 3G Universal Mobile Telecommunications System (UMTS) by the 3rd Generation Partnership Project (3GPP), operates over 5 MHz channels with variable spreading factors up to 512, supporting data rates up to 2 Mbps while incorporating advanced power control to address the near-far issue in higher-mobility scenarios.[44]Spread Spectrum Multiple Access and Non-Orthogonal Multiple Access
Spread spectrum multiple access techniques extend code-division multiple access (CDMA) principles by employing wideband transmission to enable multiple users to share the channel while providing resistance to interference and jamming. These methods spread the signal across a broader bandwidth than necessary for the information rate, allowing user separation through distinct spreading patterns and enhancing robustness in noisy environments.[45] Frequency-hopping spread spectrum (FHSS) achieves multiple access by assigning unique hopping sequences to each user, where the transmitter and receiver rapidly switch carrier frequencies according to a pseudorandom pattern synchronized between them. This orthogonal-like separation via time-varying frequency slots minimizes inter-user interference, while the wideband nature disperses narrowband jamming or interference over the spectrum, improving the signal-to-interference ratio through despreading at the receiver. A representative implementation is in Bluetooth, which uses FHSS with 79 channels in the 2.4 GHz band, hopping 1600 times per second to support multiple piconets while mitigating coexistence issues with other systems.[45][46] Direct-sequence spread spectrum (DSSS) for multiple access spreads each user's signal using a unique pseudonoise code sequence, modulating the data onto a higher-rate chip stream to achieve wideband transmission. User separation occurs via code orthogonality or correlation properties, with the receiver despreading only the intended signal to collapse it back to the original bandwidth, rejecting others as noise; this provides anti-interference benefits by processing gain, where the signal power concentrates while interference spreads. Although baseline DSSS is covered in CDMA, its spread spectrum extensions emphasize hybrid or advanced coding for enhanced multi-user capacity in ad hoc networks. Non-orthogonal multiple access (NOMA) represents a paradigm shift by intentionally allowing signals from multiple users to overlap in time, frequency, and code domains, relying on advanced receiver processing to distinguish them rather than orthogonal resource allocation. This enables higher spectral efficiency compared to orthogonal multiple access (OMA) techniques, particularly in scenarios requiring massive connectivity, such as IoT deployments, by supporting more users per resource block without proportional bandwidth expansion.[47][48] In power-domain NOMA, superposition coding combines user signals at the transmitter by allocating unequal power levels based on channel gains, with weaker users receiving higher power to ensure decodability. The received signal for two users is modeled as y = h_1 \sqrt{P_1} s_1 + h_2 \sqrt{P_2} s_2 + n, where h_i denotes the channel coefficient, P_i the allocated power, s_i the unit-power symbol, and n the additive noise. At the receiver, successive interference cancellation (SIC) decodes the stronger signal first by treating the weaker as noise, subtracts it, and then decodes the weaker signal; SIC error rates depend on power allocation and channel estimation accuracy, with propagation errors mitigated by ordering users by descending channel gain. The achievable rate region satisfies R_1 + R_2 \leq \log_2 \left(1 + \frac{P_1 + P_2}{N_0}\right), where N_0 is the noise power, demonstrating NOMA's capacity advantage over OMA's halved sum rate in shared resources.[49][47][48] Code-domain NOMA spreads user signals using low-density or sparse codebooks, allowing partial overlap and joint detection via message-passing algorithms, which reduces multi-user detection complexity compared to traditional CDMA. This variant enhances massive connectivity by accommodating more users with overlapping codes, leveraging sparsity to lower interference.[48][47] Key variants include multi-user shared access (MUSA), a code-domain scheme using complex spreading sequences and SIC for uplink grant-free access, supporting more users than OFDMA in overload scenarios. Pattern-division multiple access (PDMA) employs non-orthogonal patterns on subcarriers, such as a 5x3 pattern matrix for user-resource mapping, enabling flexible connectivity with reduced detection complexity via belief propagation.[48][47] NOMA has been evaluated in 5G New Radio (NR) systems since 2019, with studies showing integration potential in multi-cell deployments for improved uplink coverage.[50] Ongoing 6G research, including the ETSI Industry Specification Group on Multiple Access Techniques (ISG MAT) formed in 2025, targets NOMA enhancements for downlink efficiency, with the initial deliverable in early draft as of late 2025 to inform 3GPP Release 20 studies.[51][24] In dense scenarios, NOMA achieves sum rate gains of up to 30% over OMA baselines like CDMA, driven by efficient interference management and resource reuse.[47]Spatial and Resource Allocation Techniques
Space-Division Multiple Access
Space-Division Multiple Access (SDMA) is a multiple access technique that exploits the spatial dimension of the wireless channel to allow multiple users to share the same time and frequency resources simultaneously by directing signals toward specific users through antenna arrays at the base station.[52] This approach enables spatial separation of user signals, effectively creating independent channels in space without requiring orthogonal allocation in time or frequency domains.[53] The core mechanism of SDMA relies on multiple-input multiple-output (MIMO) systems, where base stations equipped with multiple antennas use beamforming to generate directional beams that focus transmitted or received signals toward particular users.[54] In MIMO configurations, spatial multiplexing is achieved by precoding signals based on channel state information (CSI), which describes the spatial characteristics of the propagation paths between the base station and users.[53] Phased array antennas adjust the phase and amplitude of signals across elements to form these beams, allowing the base station to null interference toward non-intended users while maximizing signal strength for the target.[55] Massive MIMO extends this by deploying a large number of antennas (often tens or hundreds), enhancing spatial resolution and supporting more users through finer beam control.[56] Mathematically, SDMA performance is modeled using the channel matrix \mathbf{H} \in \mathbb{C}^{K \times M}, where K is the number of users, M is the number of transmit antennas, and each row \mathbf{h}_k represents the channel vector for user k. Beamforming vectors \mathbf{w}_k are designed to precode the signal for user k, often via zero-forcing or minimum mean square error methods to minimize inter-user interference.[54] The achievable sum capacity for the multi-user downlink in SDMA simplifies to C = \log_2 \det \left( \mathbf{I} + \frac{\mathbf{H} \mathbf{Q} \mathbf{H}^*}{\sigma^2} \right), where \mathbf{Q} is the covariance matrix of the transmitted signals (diagonal for independent streams), \mathbf{H}^* is the Hermitian transpose, \mathbf{I} is the identity matrix, and \sigma^2 is the noise variance; this expression highlights how spatial separation increases capacity by leveraging the eigenvalues of the effective channel.[54] Variants of SDMA include beamforming SDMA, which dynamically adjusts beams based on real-time CSI for individual users, and sectorization, where base stations divide coverage into fixed angular sectors using directional antennas to reduce interference within each sector.[57] These variants enhance spatial reuse but require accurate user location or angle-of-arrival estimates.[58] A key advantage of SDMA is its ability to reuse time and frequency resources across spatially separated users, leading to significantly higher capacity in dense urban areas compared to traditional orthogonal methods.[59] However, it incurs substantial CSI acquisition overhead, as frequent channel estimation and feedback are needed to combat imperfect knowledge, which can degrade performance under mobility.[60] Additionally, the hardware complexity of large antenna arrays and phase shifters increases deployment costs and power consumption.[61] Early implementations of SDMA appeared in WiMAX systems during the 2000s, utilizing array antennas for adaptive beamforming to support multi-user spatial multiplexing in IEEE 802.16 networks.[62] In modern 5G networks, massive MIMO has become a cornerstone of SDMA since its standardization around 2019, with base stations employing hundreds of antennas to serve dozens of users concurrently via precise beamforming.[63] SDMA provides spatial degrees of freedom up to the number of antennas at the base station, enabling parallel data streams without resource partitioning.[52] For instance, a 64-element antenna array can support up to 16 simultaneous beams in sub-array configurations, achieving substantial multiplexing gains in millimeter-wave bands.[64]Power-Division Multiple Access
Power-division multiple access (PDMA) enables multiple users to simultaneously access the same time-frequency-code resources by assigning distinct power levels to their signals, distinguishing them in the power domain rather than through orthogonal separation. At the transmitter, superposition coding overlays the user signals, with higher power allocated to users experiencing weaker channel conditions to ensure equitable decoding opportunities. The receiver employs ordered successive interference cancellation (SIC), decoding the strongest signal first—typically from the user with the best effective channel gain—subtracting it from the composite signal, and proceeding iteratively to weaker signals. This process relies on accurate channel state information (CSI) for ordering and cancellation.[65][66] To promote fairness among users with heterogeneous channel gains or quality-of-service demands, power allocation algorithms dynamically adjust the power shares P_k for each user k, often balancing individual rates against total throughput. Common approaches include water-filling adaptations or cognitive radio-inspired methods that prioritize edge users. PDMA operates in both uplink and downlink configurations: in the downlink, the base station controls superposition and power levels; in the uplink, mobile users coordinate transmit powers under base station guidance to achieve desired received power disparities. It is frequently integrated with non-orthogonal multiple access (NOMA) frameworks, leveraging power-domain multiplexing alongside other domains for enhanced connectivity in dense networks.[65][67] PDMA enhances spectral efficiency in overloaded scenarios where the number of users exceeds available orthogonal resources, allowing up to 50% more users per cell compared to traditional orthogonal methods by exploiting multi-user interference constructively via SIC. However, drawbacks include error propagation during SIC—if a strong signal is decoded incorrectly, it contaminates subsequent decodings—and the need for stringent power control to maintain sufficient separation (typically 3-6 dB) amid channel variations. These challenges necessitate robust CSI feedback and error-correcting codes to mitigate performance degradation.[66][68] The achievable rate for the k-th user, ordered by decreasing channel gain |h_1|^2 \geq |h_2|^2 \geq \cdots \geq |h_K|^2, is expressed as R_k = \log_2 \left( 1 + \frac{ |h_k|^2 P_k }{ \sum_{j=k+1}^K |h_j|^2 P_j + N_0 } \right), where P_k denotes the allocated power for user k, and N_0 is the noise power spectral density. The weak users treat strong user signals as noise, while strong users perform full SIC. Optimization of \{P_k\} subject to total power constraint \sum P_k \leq P_{\max} maximizes the sum rate \sum R_k, solvable via convex programming or greedy algorithms.[69][66] Initial PDMA concepts emerged in the late 1990s, with Mazzini's 1998 proposal demonstrating its viability for asynchronous environments through power diversity. Subsequent developments integrated PDMA principles into 5G research, with enhancements in 3GPP Release 15 and later for uplink scenarios supporting massive connectivity, though full standardization remains in study items for improved power-domain superposition in non-standalone deployments.[65][70] Performance evaluations highlight PDMA's power efficiency; relative to equal-power allocation, it provides coding gains of 2-4 dB in bit error rate at moderate signal-to-noise ratios. In a representative two-user downlink case with 3 dB power separation (e.g., 70% power to the weak user), PDMA achieves approximately 1.5 times the sum capacity of orthogonal frequency-division multiple access under Rayleigh fading channels with SNR=20 dB.[66][71]Contention-Based and Packet-Oriented Methods
Carrier Sense Multiple Access Protocols
Carrier Sense Multiple Access (CSMA) protocols enable multiple devices to share a communication channel in packet networks by requiring stations to sense the carrier before transmission, thereby minimizing collisions through decentralized contention resolution. Developed primarily for radio and local networks, CSMA improves upon blind random access by deferring transmissions when the channel is detected as busy, allowing for higher efficiency in low-to-moderate load scenarios.[72] The core mechanism of CSMA involves a station monitoring the channel: if idle for a minimum duration, it proceeds with transmission; otherwise, it postpones the attempt according to the protocol variant. Upon collision, stations employ backoff algorithms, typically binary exponential backoff, where the contention window size doubles after each failure, selecting a random slot within it for retry. This process ensures progressive deferral to resolve conflicts fairly. In wired environments, CSMA/CD extends this by actively detecting collisions during transmission via signal monitoring, aborting and jamming the channel upon detection before backing off. Conversely, in wireless settings, CSMA/CA relies on avoidance techniques like interframe spacing and optional RTS/CTS handshakes, where a sender requests channel clearance and the receiver broadcasts approval, updating the network allocation vector (NAV) to silence potential interferers.[72][73][74] CSMA variants differ in persistence strategies to balance aggressiveness and collision risk. 1-persistent CSMA transmits immediately (probability 1) if the channel is idle, or continuously monitors and transmits upon idleness, risking high collisions from synchronized arrivals. Non-persistent CSMA defers for a random time if busy, reducing but not eliminating overlaps. p-persistent CSMA hybridizes by transmitting with probability p if idle and deferring with 1-p, optimizing for slotted channels. These were formalized and analyzed for radio channels, showing p-persistent yielding the highest capacity under tuned p.[72] Key advantages of CSMA include its fully decentralized nature, requiring no central arbitrator, and superior performance at low loads where sensing prevents most collisions. However, disadvantages arise in wireless deployments: the hidden terminal problem occurs when nodes cannot sense each other but interfere at a common receiver, while the exposed terminal problem unnecessarily silences nodes due to unrelated transmissions; additionally, fairness issues emerge as backoff dynamics can favor nodes with fewer retries, leading to capture effects.[72][75][76] Throughput analysis provides insight into CSMA's efficiency. For baseline pure ALOHA without sensing, throughput is given byS = G e^{-2G},
maximizing at 18.4% when offered load G = 0.5. CSMA approximations improve this; one model yields
S \approx \frac{1}{1 + 6.5 e^{-\lambda}},
where λ is the normalized attempt rate, approaching 81% as λ increases under negligible propagation delay. Detailed derivations for non-persistent CSMA give
S = \frac{G e^{-aG}}{G(1 + 2a) + e^{-aG}},
with a as the propagation delay ratio, achieving ~81% capacity for small a. All variants show instability at high loads without stabilization controls.[72] Prominent implementations trace to Ethernet, which pioneered CSMA/CD in the 1970s for 10 Mbps shared-bus local networks, evolving into modern switched variants. Wi-Fi, standardized as IEEE 802.11 in 1997, adopts CSMA/CA with RTS/CTS for collision-prone wireless links, supporting rates up to multi-Gbps in later amendments.[73][74] Under typical conditions, CSMA/CA attains maximum throughput of approximately 80% of channel capacity, though this degrades with overheads like ACKs and beacons. In saturation, collision probability nears 1, causing exponential backoff to dominate and throughput to plummet, highlighting the need for load management.[72][74]