Electronics Guide

Principles of Spectrum Spreading

An ordinary communication signal occupies a bandwidth roughly proportional to its data rate. Spread-spectrum systems break this proportionality by modulating the data with a spreading code whose rate, measured in chips per second, greatly exceeds the data rate. The result is a transmitted signal whose bandwidth is set by the chip rate rather than the data rate, often hundreds or thousands of times wider than a conventional signal carrying the same information. The coarse civil signal of the Global Positioning System illustrates the extreme: a navigation message of 50 bits per second occupies roughly 2 MHz of spectrum because the spreading code runs at 1.023 million chips per second.

Two conditions define a true spread-spectrum system. First, the transmitted bandwidth must be much greater than the bandwidth of the information. Second, the spreading must be governed by a code independent of the data, known to the intended receiver but appearing random to others. These conditions distinguish spread spectrum from wideband schemes, such as wideband frequency modulation, that expand bandwidth as a byproduct of the modulation itself rather than through an independent code.

The benefit of spreading emerges at the receiver. Correlating the received signal with a synchronized replica of the spreading code despreads the wanted signal, restoring it to its narrow information bandwidth, while simultaneously spreading any narrowband interference across the wide band. A narrowband filter following the despreader then passes most of the wanted signal energy but only a small fraction of the spread interference, producing a gain in signal-to-interference ratio that is the central advantage of the technique.

Information theory frames the trade explicitly. The Shannon–Hartley relation shows that a given capacity may be purchased either with high signal-to-noise ratio in a narrow band or with low signal-to-noise ratio in a wide one. Spread spectrum chooses the second path, accepting a signal that may sit at or below the noise floor in exchange for bandwidth. That choice is what makes covert operation, interference tolerance, and code-division sharing possible in the first place.

Historical Development

The idea of hopping a carrier to frustrate jamming was patented during the Second World War. On August 11, 1942, the United States Patent Office granted patent 2,292,387, "Secret Communication System," to the actress Hedy Lamarr, filing under her married name, and the composer George Antheil. Their scheme synchronized transmitter and receiver with slotted paper rolls borrowed from player-piano mechanisms so that a radio link controlling a torpedo would step through a fixed sequence of frequencies. The Navy did not deploy the invention, and the patent expired before spread spectrum became practical, but the concept it described is precisely modern frequency hopping.

Practical systems awaited electronics capable of generating and synchronizing long pseudorandom codes. Classified military programs of the 1950s and 1960s developed direct-sequence and hopping links for anti-jam and covert communication, and the Global Positioning System, whose first satellites launched in 1978, made direct-sequence spreading the foundation of satellite navigation. Commercial adoption followed regulatory change: in 1985 the Federal Communications Commission authorized unlicensed spread-spectrum operation in the industrial, scientific, and medical bands, opening the path to cordless telephones, wireless local area networks, and, eventually, to the code-division cellular systems of the 1990s.

Direct-Sequence Spread Spectrum

Direct-sequence spread spectrum (DSSS) spreads the signal by multiplying the data stream directly with a high-rate pseudonoise code. Each data bit is represented by a fixed sequence of code chips, so the transmitted waveform changes at the chip rate. The wide bandwidth follows directly from the rapid chip transitions, and the structure of the code determines the spectral and correlation properties of the signal. The ratio of chip rate to symbol rate, called the spreading factor, sets how many chips carry each symbol and therefore how much bandwidth the signal consumes.

Spreading and Despreading

At the transmitter, the data signal and the spreading code, both represented as sequences of plus and minus one, multiply together before modulating the carrier. Because the code chips change far faster than the data bits, the product occupies the wide bandwidth of the code. At the receiver, multiplication by a time-aligned copy of the same code reverses the operation. Since the product of the code with itself equals one at every chip, the wanted data emerge intact, restored to their original bandwidth. Any signal not carrying the matching code, including interference and the signals of other users, is instead multiplied by the code for the first time at the receiver and spread across the wide band.

Because the chip stream is a rectangular pulse train, the transmitted power spectrum follows a squared sinc envelope whose main lobe is twice the chip rate wide for binary phase-shift keying. A pulse-shaping filter, typically a root-raised-cosine response, usually replaces the rectangular chip in practical transmitters to contain the spectrum within a regulatory mask. The despreading correlator is matched to the transmitted chip shape so that the shaping costs no processing gain.

Representative Chip Rates

Real systems span several decades of chip rate. IEEE 802.11b runs at 11 million chips per second and occupies roughly 22 MHz; its 1 and 2 Mbit/s modes spread each symbol with an 11-chip Barker sequence, while the 5.5 and 11 Mbit/s modes substitute complementary code keying, which carries several bits in the choice of code word and so trades processing gain for throughput. The IS-95 cellular air interface uses 1.2288 million chips per second in a 1.25 MHz carrier, and its Universal Mobile Telecommunications System successor uses 3.84 million chips per second in a 5 MHz carrier. IEEE 802.15.4, the physical layer beneath Zigbee, maps each four-bit symbol to a 32-chip sequence at 2 million chips per second to deliver 250 kbit/s in the 2.4 GHz band.

Resistance to Narrowband Interference

The despreading operation reveals why direct-sequence systems resist narrowband interference. A strong interferer that occupies a small part of the band passes through the receiver's despreading multiplier and emerges spread across the full code bandwidth, with its power density correspondingly reduced. The narrowband filter after despreading rejects most of this spread energy. The wanted signal, by contrast, is concentrated by despreading and passes through the filter with little loss. The receiver thereby suppresses interference in proportion to the ratio of the spread bandwidth to the information bandwidth.

Multipath Behavior

The sharp autocorrelation of a good spreading code allows a direct-sequence receiver to distinguish signal replicas that arrive by different paths and that are delayed by more than one chip period. A rake receiver exploits this property by assigning separate correlators, or fingers, to the strongest delayed copies and combining them constructively, usually with maximal-ratio weighting. Rather than suffering from multipath, such a receiver gains diversity from it, which is a notable advantage in mobile environments. The resolution is set by the chip period: a 3.84 Mchip/s signal resolves paths separated by about 260 nanoseconds, or roughly 78 meters of extra propagation distance, so rake combining pays off in outdoor cells with long delay spreads and offers less in small rooms where all echoes fall inside one chip.

Frequency-Hopping Spread Spectrum

Frequency-hopping spread spectrum (FHSS) spreads the signal in a different manner. Instead of widening the instantaneous bandwidth, it transmits a relatively narrowband signal whose carrier frequency changes rapidly according to a pseudorandom hopping pattern. Over time the signal visits many frequencies across a wide band, so the average occupied bandwidth is large even though the bandwidth at any instant is modest. The hopping sequence, derived from a pseudonoise code, is known to the intended receiver, which retunes in step to follow the signal. Because each hop is narrowband, frequency hopping tolerates cheaper filters and less linear amplifiers than direct sequence, which helps explain its popularity in low-cost consumer radios.

Fast and Slow Hopping

Frequency-hopping systems are classified by the relationship between the hop rate and the symbol rate. In slow hopping, several symbols are transmitted on each frequency before the next hop. In fast hopping, the carrier hops several times within a single symbol, which provides additional protection because an interferer must corrupt every hop of a symbol to destroy it. The choice involves a trade-off between the complexity and stability of the rapidly retuning synthesizer and the degree of robustness obtained. Fast hopping also forbids coherent demodulation across a symbol, so such systems typically use noncoherent frequency-shift keying and accept the accompanying loss in sensitivity.

Bluetooth as a Worked Example

Bluetooth basic rate and enhanced data rate operation illustrates slow hopping in the 2.4 GHz band. The radio divides the band into 79 channels spaced 1 MHz apart and hops nominally 1600 times per second, giving a 625-microsecond time slot on each channel. Adaptive frequency hopping, introduced in Bluetooth 1.2 in 2003, improves coexistence by measuring channel quality and removing persistently occupied channels from the hopping set, so that a Bluetooth piconet steers around a wireless local area network rather than colliding with it repeatedly. Bluetooth Low Energy uses a different channel plan of 40 channels spaced 2 MHz apart, of which 37 carry connection data and 3 serve as primary advertising channels, and applies its own adaptive hopping across the data channels.

Interference and Jamming Behavior

A frequency-hopping signal avoids a narrowband interferer or a jammer by spending only a fraction of its time on any single frequency. When the signal happens to hop onto an occupied frequency, the affected symbols may be corrupted, but error-correcting codes and interleaving readily recover from such intermittent losses. Against a partial-band jammer that concentrates its power on a portion of the band, frequency hopping forces the jammer to choose between covering the whole band weakly or part of it strongly, and coding defeats either choice. Without coding, partial-band jamming is in fact the jammer's best strategy against a hopper, because concentrating power produces occasional deep hits whose errors dominate performance; interleaved forward error correction removes that advantage by spreading each code word across many hops. This graceful behavior in hostile environments accounts for the prevalence of frequency hopping in tactical military radios and in the adaptive hopping used by short-range commercial systems to coexist with other devices.

Processing Gain and Jamming Margin

Processing gain quantifies the advantage that spreading confers against interference. It equals the ratio of the spread bandwidth to the information bandwidth, or equivalently the ratio of the chip rate to the data rate in a direct-sequence system. Expressed in decibels, the processing gain is ten times the base-ten logarithm of this ratio. A system that spreads a signal by a factor of one thousand achieves a processing gain of 30 dB, meaning that despreading improves the signal-to-interference ratio by that amount relative to the ratio present in the wide channel. The 11-chip Barker sequence of IEEE 802.11b yields about 10.4 dB, a modest figure adequate for an unlicensed indoor band, while the GPS coarse acquisition signal, spreading 50 bits per second to 1.023 million chips per second, yields about 43 dB, which is what allows a navigation signal arriving well below the thermal noise floor to be recovered at all.

Processing gain does not improve performance against broadband thermal noise, because such noise already fills the channel and is unaffected by spreading. Its benefit applies to interference that the despreader spreads, including narrowband jammers, the signals of other spread-spectrum users sharing the band, and intentional jamming. Spreading likewise does not change the energy required per bit for a given error rate; it changes only how that energy is distributed in frequency.

Jamming margin expresses how much stronger than the wanted signal an interferer may be while the receiver still operates. It equals the processing gain reduced by the signal-to-noise ratio the demodulator requires and by implementation losses in the receiver, so a link with 43 dB of processing gain, a 7 dB demodulation requirement, and 2 dB of implementation loss tolerates interference roughly 34 dB above the wanted signal. A large processing gain therefore translates directly into the capacity to operate beneath strong interference, which is the defining capability of spread spectrum. The margin is not unlimited: front-end saturation, analog-to-digital converter dynamic range, and code-tracking breakdown often cap usable margin below the theoretical figure.

Pseudonoise Sequences

The spreading code is the heart of any spread-spectrum system, and its properties determine how well the system performs. Such codes are called pseudonoise or pseudorandom sequences because they appear noiselike and statistically random, yet they are generated deterministically and can be reproduced exactly by an authorized receiver.

Desired Properties

A good spreading code possesses three properties. It has a sharp autocorrelation, meaning that the code correlates strongly with itself only when perfectly aligned and weakly at all other shifts, which enables timing acquisition and multipath resolution. It has low cross-correlation with the codes assigned to other users, which limits the mutual interference among signals sharing a band. And it has balance and randomness in the distribution of its chips, which produces a smooth, noiselike spectrum. These goals conflict: a family cannot simultaneously maximize its size and drive all cross-correlations to zero, and bounds such as the Welch bound fix the best trade available for a given code length and family size.

Common Code Families

Maximal-length sequences, generated by an n-stage linear feedback shift register with a primitive feedback polynomial, run for 2ⁿ − 1 chips before repeating and achieve nearly ideal autocorrelation, with every off-peak value equal to −1. They are trivial to produce in hardware, but the number of distinct sequences of a given register length is limited and their mutual cross-correlation is not always small. Gold codes, formed by modulo-two addition of a preferred pair of maximal-length sequences at every relative shift, provide a family of 2ⁿ + 1 codes of length 2ⁿ − 1 with bounded cross-correlation, which makes them well suited to systems in which many users transmit simultaneously. The Global Positioning System uses Gold codes of 1023 chips, built from two 10-stage registers and repeating every millisecond, to give each satellite a distinct signature on a shared frequency. Kasami sequences offer similarly favorable cross-correlation for large code sets. Walsh functions, which are mutually orthogonal when perfectly synchronized, separate channels in synchronous systems: IS-95 assigns 64-chip Walsh functions on the forward link, and the Universal Mobile Telecommunications System generalizes them into orthogonal variable spreading factor codes, whose tree structure lets one carrier serve users at different rates while keeping their codes orthogonal.

Codes and Security

Statistical randomness is not cryptographic strength. A maximal-length sequence is a linear recursion, so observing 2n consecutive chips is enough to recover the generator polynomial and predict the rest, a computation the Berlekamp–Massey algorithm performs efficiently. Systems that rely on the spreading code for security therefore drive the code generator from a cryptographic algorithm rather than a bare shift register, as the encrypted P(Y) code of the military GPS signal does. Civil systems make no such claim: their codes are published so that any receiver may use them, and confidentiality, if required, comes from encrypting the message itself.

Code-Division Multiple Access

Spread spectrum enables a powerful method of sharing a channel among many users. In code-division multiple access (CDMA), every user transmits over the same frequency band at the same time, distinguished not by frequency or time slot but by a unique spreading code. A receiver despreads the signal with the code of the desired user, concentrating that user's energy while leaving the other users spread as noiselike interference. The low cross-correlation among the assigned codes keeps this multiple-access interference manageable, so that many users coexist in the same band.

Capacity and Interference

The capacity of a CDMA system is limited not by a fixed number of channels but by the aggregate interference that all active users present to one another. Each additional user raises the interference floor slightly, so capacity degrades gracefully as load increases rather than collapsing at a hard limit. Operators exploit that softness: because speech is active only part of the time, suppressing transmission during silence reduces average interference and raises the number of conversations a carrier supports, and sectorized antennas divide the interference further. The same softness has a cost, since every cell's load raises its neighbors' noise floor, and an overloaded cell shrinks its own coverage, a phenomenon operators call cell breathing.

Power Control and the Near-Far Problem

Interference-limited operation makes power control essential. A handset close to a base station would otherwise swamp the despread signal of a distant one, because the despreader suppresses the unwanted signal only by the processing gain, and the path-loss difference between a near and a far user can easily exceed it. Practical systems therefore combine an open-loop estimate, in which the handset infers path loss from the received forward-link power, with a fast closed loop in which the base station commands power up or down. IS-95 issues those commands 800 times per second, and wideband CDMA raises the rate to 1500 times per second, one command per slot, to track fast fading. The loops span a dynamic range of many tens of decibels. Soft handover, in which a handset communicates with several base stations at once and their signals are combined, is a further consequence of universal frequency reuse and has no counterpart in systems that assign distinct frequencies to adjacent cells.

Role in Cellular Systems

Code-division multiple access shaped a generation of cellular technology. Second-generation systems based on the IS-95 standard and the third-generation Universal Mobile Telecommunications System employed wideband CDMA as their air interface, exploiting soft capacity, frequency reuse in every cell, and rake reception of multipath. Fourth- and fifth-generation networks moved their main traffic channels to orthogonal frequency-division multiplexing, which handles wide bandwidths and multipath with simpler equalization, but spreading did not disappear. Constant-envelope sequences with sharp correlation properties still carry random-access preambles, reference signals, and synchronization signals, where the ability to detect a signal of unknown timing is precisely what is needed, and satellite links continue to rely on code division.

Jam Resistance and Low Probability of Interception

The military origins of spread spectrum reflect two properties that remain among its most valued. Jam resistance follows from processing gain. An adversary attempting to deny communication must inject enough power into the wide spread bandwidth to overcome the wanted signal after despreading, and the processing gain forces the jammer to expend far more power than a conventional signal would require. Frequency hopping compounds the difficulty by denying the jammer knowledge of where the signal will appear next, while direct-sequence spreading buries the signal beneath a wideband waveform that resists narrowband jamming.

Low probability of interception (LPI) and low probability of detection (LPD) follow from spreading the signal energy thinly across a wide band. Because the transmitted power is distributed over a bandwidth far larger than the information requires, the power spectral density of the signal can fall below the noise floor of an unintended receiver, rendering the transmission difficult to detect, let alone demodulate, without knowledge of the spreading code. A direct-sequence signal can be made to appear as a slight rise in the background noise, and a frequency-hopping signal presents only brief, scattered bursts to an observer who does not know the hopping pattern.

These advantages are relative rather than absolute. An energy detector integrating over the whole band, a wideband receiver searching for the periodicity that a repeating short code imposes on the signal, or a direction finder observing hop bursts can all reveal a transmission that no single narrowband receiver would notice, which is why covert systems use long, cryptographically generated codes and limit transmitter power and duty cycle. The secrecy of the spreading code likewise provides a layer of access control but does not substitute for cryptographic protection of the message.

Synchronization

A spread-spectrum receiver cannot recover any information until it aligns its local code with the incoming signal, because despreading requires the local and received codes to match in both code phase and, for frequency hopping, hop timing. Synchronization is therefore a critical and often demanding function, conventionally divided into acquisition and tracking.

Acquisition

Acquisition is the initial process of bringing the local code into coarse alignment with the received code, typically within a fraction of a chip. Because the receiver does not know the correct code phase in advance, it must search across the possible alignments until correlation indicates a match. The search is two-dimensional, since Doppler shift and oscillator error leave the carrier frequency uncertain as well, and each candidate cell must be dwelt upon long enough for the correlation to rise above the noise. A serial search tests candidate phases one at a time and is simple but slow; parallel correlator banks and transform-based methods, which compute all code phases at once by multiplying spectra and applying an inverse fast Fourier transform, acquire far more quickly at the cost of memory and arithmetic. Practical receivers shrink the search space with outside information: assisted GNSS receivers obtain approximate time, position, and satellite ephemeris over a network link, and many systems transmit a short unmodulated preamble or a separate pilot channel that a receiver may acquire before attempting the traffic channel.

Tracking

Once acquisition achieves coarse alignment, tracking maintains fine synchronization as the relative timing drifts because of clock differences and motion. A delay-locked loop is the usual mechanism, comparing correlations of the received signal with slightly early and late copies of the code and driving the difference to zero so that the punctual correlator stays centered on the incoming signal. Narrowing the early-late spacing sharpens the discriminator and reduces the tracking error caused by multipath, at the cost of a smaller pull-in range. A parallel carrier-tracking loop, typically a Costas loop that is insensitive to data modulation, maintains frequency and phase alignment. The precision of code tracking is what allows spread-spectrum systems to measure propagation delay accurately: tracking to a small fraction of a chip is what turns the 293-meter chip of the GPS coarse acquisition code into a pseudorange good to a few meters, and carrier-phase processing refines it further.

Hybrid and Related Spreading Techniques

Direct sequence and frequency hopping are not exclusive. Hybrid systems spread each hop with a direct-sequence code, combining the instantaneous interference rejection of one method with the frequency agility of the other and easing the near-far problem that afflicts pure direct-sequence networks without power control. Military tactical waveforms commonly take this form. Time hopping, in which transmission occurs in pseudorandomly selected slots, provides a third axis and appears in impulse-radio ultra-wideband systems.

Chirp spread spectrum spreads by sweeping the carrier linearly across a wide band rather than by applying a code. The swept pulse compresses in a matched filter to a narrow peak, delivering processing gain and precise time resolution while remaining tolerant of frequency offset, which is why chirp radar has used the technique for decades. LoRa applies the same principle to low-power wide-area networking, selecting among spreading factors that trade data rate for link budget, and IEEE 802.15.4 defines a chirp physical layer as well. Impulse ultra-wideband takes the opposite route to a wide spectrum, radiating subnanosecond pulses whose bandwidth follows from their brevity; regulators class it separately from coded spread spectrum, though its ranging applications overlap closely.

Implementation Considerations

Building a spread-spectrum radio imposes demands that a narrowband design avoids. The correlator dominates digital complexity, since the receiver must multiply and accumulate at the chip rate across every code phase under test; parallel acquisition engines and frequency-domain correlation trade silicon area and memory for acquisition time, a trade that dictates how quickly a satellite navigation receiver reports a first fix. Reference-oscillator stability governs how large the frequency search must be and how long a coherent integration may run, so receivers designed for weak signals depend on temperature-compensated or oven-controlled oscillators, or on frequency aiding from an external source.

The analog front end must accommodate interference that the digital section is designed to reject. Because the wanted signal may lie far below a nearby interferer, the low-noise amplifier and mixer must remain linear over a wide dynamic range, and the analog-to-digital converter needs enough bits of headroom that a strong in-band interferer does not clip the signal before despreading can suppress it. Automatic gain control must respond quickly enough to track bursts without stealing the headroom the correlator needs. Frequency-hopping designs shift the burden to the synthesizer, which must settle to the next channel within the guard interval between hops; direct digital synthesis or fast-switching fractional-N architectures serve where a conventional phase-locked loop settles too slowly.

System-level choices follow from these constraints. Longer codes improve cross-correlation and security but lengthen acquisition; shorter codes acquire quickly but repeat often enough to be detected and to correlate poorly with one another. Higher chip rates buy processing gain and ranging resolution but raise sampling rates, power consumption, and cost. Regulatory limits on emitted power and spectral density then bound the achievable link budget, so the designer settles the chip rate, code length, coding scheme, and antenna gain together rather than in isolation.

Regulatory Context

Unlicensed spread-spectrum operation is shaped by the rules that permit it. In the United States, Part 15 of the Federal Communications Commission's regulations governs the 902–928 MHz, 2.4 GHz, and 5.725–5.850 GHz bands. Section 15.247 once required direct-sequence systems to demonstrate at least 10 dB of processing gain, a rule that had the effect of mandating spreading rather than merely permitting it. A 2002 order removed that requirement and placed direct-sequence and other digital modulation schemes under common limits, principally a minimum 6 dB bandwidth of 500 kHz and a maximum power spectral density of 8 dBm in any 3 kHz band. Modern wireless local area network and personal-area network radios consequently use whatever modulation suits them, and spreading survives where it earns its keep on technical merit rather than by regulation.

Frequency-hopping systems remain subject to their own conditions, including minimum numbers of hopping channels, limits on average dwell time per channel, and requirements that channels be used approximately equally. European rules administered through harmonized standards impose comparable constraints, adding duty-cycle limits in some sub-gigahertz bands and requiring listen-before-talk or adaptive behavior in others. Designers targeting several regions must satisfy the intersection of these regimes, which often decides the hop rate, channel plan, and duty cycle of a product before any link-budget consideration does.

Applications

Spread-spectrum techniques appear across a broad range of systems, civil and military, that exploit one or more of their distinctive properties.

Satellite Navigation

Global navigation satellite systems, including the Global Positioning System, rely on direct-sequence spread spectrum. Each satellite broadcasts a unique spreading code, and a receiver measures its position by correlating against these codes to determine the propagation delay, and hence the distance, from several satellites at once. The sharp autocorrelation of the codes provides the precise timing that positioning demands, and code division allows all satellites to share the same frequency. Galileo and BeiDou follow the same code-division approach, whereas the legacy GLONASS signals separated satellites by frequency instead, an arrangement later augmented with code-division signals. Modernized navigation signals add binary offset carrier modulations that move signal energy away from the band center, improving code-tracking accuracy and easing coexistence among the constellations.

Cellular and Wireless Systems

Code-division multiple access served as the foundation of major second- and third-generation cellular networks, and spreading concepts persist in the synchronization and random-access channels of later generations. Short-range wireless technologies apply spread spectrum widely. Adaptive frequency hopping allows Bluetooth piconets to coexist in the crowded 2.4 GHz band, direct-sequence spreading carried the early IEEE 802.11 and 802.11b wireless local area networks and still underlies IEEE 802.15.4 radios, and chirp spreading gives LoRa the link budget it needs for kilometers of range at milliwatt power levels. These commercial uses exploit spread spectrum chiefly for its resistance to interference and its ability to share unlicensed spectrum gracefully.

Military and Specialized Links

Military communication systems employ spread spectrum for jam resistance and for low probability of interception and detection, combining frequency hopping and direct-sequence spreading in tactical radios and secure data links, and pairing them with cryptographic key management that supplies the code sequences. Beyond communication, the precise ranging that spread spectrum permits supports radar, distance-measurement equipment, and secure ranging in keyless entry systems, where a coded waveform resists the relay attacks that defeat simple signal-strength methods. Deep-space missions apply pseudonoise ranging codes on the same principle to measure the distance to a spacecraft across hundreds of millions of kilometers, and the technique finds use in any application that benefits from operating beneath interference or beneath the notice of an unintended observer.

Summary

Spread-spectrum communication trades bandwidth for robustness by spreading a signal with a pseudonoise code independent of the data and despreading it with a synchronized replica at the receiver. Direct-sequence systems spread by multiplying the data with a high-rate code, widening the instantaneous bandwidth and resisting narrowband interference, while frequency-hopping systems spread by retuning a narrowband carrier across many frequencies in a pseudorandom pattern, evading jammers and interference. Processing gain, the ratio of spread to information bandwidth, measures the resulting advantage against interference and sets the jamming margin. Pseudonoise sequences with sharp autocorrelation and low cross-correlation, such as maximal-length, Gold, and Walsh codes, make spreading and code-division access possible, though linear sequences offer no cryptographic protection on their own. Code-division multiple access lets many users share a band, limited by mutual interference and dependent on fast power control. Spread spectrum confers jam resistance and a low probability of interception and detection, demands careful acquisition and tracking for synchronization, and underlies satellite navigation, cellular and short-range wireless systems, and secure military links. Even where newer air interfaces have moved traffic to orthogonal frequency-division multiplexing, spreading endures wherever a receiver must find a signal of unknown timing, measure delay precisely, or work beneath the noise.

Related Topics