Electronics Guide

Defining Spectrum Efficiency

Spectrum efficiency can be defined from multiple perspectives:

Technical efficiency: The amount of information (bits) transmitted per unit bandwidth (hertz) and time. This is the classic Shannon-inspired view that focuses on modulation and coding efficiency.

Spatial efficiency: How effectively spectrum is reused across geographic area. Cellular systems achieve high spatial efficiency through frequency reuse in small cells.

Temporal efficiency: The fraction of time that spectrum is actually carrying useful traffic. Low temporal efficiency indicates spectrum that is allocated but often idle.

Economic efficiency: The value generated per unit of spectrum. This broader view considers not just technical throughput but the economic and social benefit derived from spectrum use.

A comprehensive approach to spectrum efficiency must address all these dimensions, recognizing that improvements in one area may involve tradeoffs in others.

Efficiency Metrics

Quantifying spectrum efficiency requires appropriate metrics:

Bits per hertz (b/Hz): The fundamental measure of spectral efficiency for a single link. Modern systems achieve 5-10 b/Hz through advanced modulation and coding.

Bits per hertz per square kilometer (b/Hz/km2): Area spectral efficiency accounts for spatial reuse. Dense small-cell deployments dramatically increase this metric.

Spectrum utilization: The percentage of time and geographic area where allocated spectrum actually carries traffic. Studies often find utilization well below 50% for many bands.

Capacity per MHz: The total system throughput divided by allocated bandwidth, sometimes normalized by area or population served.

Different metrics are appropriate for different contexts. Broadcast services prioritize coverage per MHz; mobile networks emphasize capacity per MHz per square kilometer; point-to-point links focus on bits per Hz on individual paths.

Sources of Inefficiency

Understanding why spectrum is used inefficiently guides the selection of improvement techniques:

Static allocation: Fixed frequency assignments cannot adapt to varying demand. Spectrum assigned to a service may be unused when that service is inactive.

Guard bands: Frequency gaps between services or operators prevent interference but carry no traffic. Legacy technology constraints often require larger guards than modern equipment needs.

Peak dimensioning: Systems sized for peak demand operate below capacity most of the time. Spectrum allocated for busy-hour traffic sits partially idle during off-peak periods.

Geographic over-allocation: Spectrum licensed for an entire region may only be used in urban areas, leaving rural spectrum unutilized.

Technology limitations: Older systems use less efficient modulation and cannot adapt to channel conditions. Equipment designed for worst-case conditions wastes capacity when conditions are good.

Cognitive Radio Concepts

A cognitive radio possesses several key capabilities:

Spectrum sensing: The ability to monitor the radio environment and detect which frequencies are occupied by other transmissions. Sensing enables the radio to identify opportunities for spectrum access.

Spectrum analysis: Beyond simple detection, analysis characterizes the signals present, including their modulation type, power level, and likely purpose. This information guides access decisions.

Spectrum decision: Based on sensing and analysis, the radio decides which frequencies to use and with what parameters. Decisions must balance opportunity against risk of interference.

Spectrum mobility: When conditions change, the radio must be able to vacate frequencies and move to alternatives. This includes responding when a primary user appears.

The cognitive radio concept extends beyond frequency selection to include adaptive modulation, power control, and protocol selection based on observed conditions.

Spectrum Sensing Techniques

Detecting the presence or absence of other signals is fundamental to cognitive radio operation. Various sensing approaches offer different tradeoffs:

Energy detection: The simplest approach measures total received power in a band and compares against a threshold. Energy detection is computationally simple but cannot distinguish signal from noise at low signal-to-noise ratios and cannot identify signal types.

Feature detection: Exploits known features of signals such as pilot tones, cyclostationary properties, or protocol-specific patterns. More sensitive than energy detection but requires knowledge of the signals to be detected.

Matched filter detection: Correlates received signal against a known waveform template. Optimal for detecting known signals but requires complete signal knowledge and is computationally intensive.

Cooperative sensing: Multiple cognitive radios share sensing information to overcome individual limitations. Spatial diversity helps with the hidden terminal problem where one radio cannot detect a signal that would interfere with another.

Cognitive Radio Architectures

Cognitive radios can be organized in various ways:

Autonomous cognitive radio: Individual radios make independent decisions based on local sensing. Simple to deploy but may lead to suboptimal collective behavior.

Networked cognitive radio: Radios coordinate through a network infrastructure, sharing sensing information and coordinating access. Enables more sophisticated spectrum sharing but requires supporting infrastructure.

Database-assisted cognitive radio: Radios query a geolocation database to determine available frequencies rather than relying solely on sensing. The database contains information about incumbent users and provides authorized channel lists based on the radio's location.

Hybrid approaches combine local sensing with database information, using sensing to verify database accuracy and detect unlisted signals.

DSA Models

Several models for dynamic access have been proposed and implemented:

Opportunistic access: Secondary users access spectrum when primary users are not using it, with no coordination. The secondary user must detect primary activity and vacate immediately when the primary returns. This model requires robust sensing to protect incumbents.

Licensed shared access (LSA): Secondary users receive formal authorization to share spectrum with incumbents under agreed conditions. Coordination mechanisms ensure protection of incumbent operations while providing more predictable access for secondary users.

Spectrum access system (SAS): A centralized system manages access to shared spectrum, assigning frequencies and parameters to users based on their location, incumbent protection requirements, and demand. The CBRS framework in the United States uses this model.

Dynamic frequency selection (DFS): In shared bands, devices must detect incumbent signals (such as radar) and avoid using frequencies where incumbents are present. This is required for 5 GHz WiFi coexistence with radar.

Implementation Challenges

Deploying dynamic spectrum access involves several challenges:

Detection reliability: Sensing must be reliable enough to protect incumbents, but overly conservative thresholds waste spectrum by avoiding frequencies that are actually available. The hidden terminal problem complicates reliable detection.

Transition timing: When primary users appear, secondary users must vacate quickly to avoid interference. But too-rapid transitions may cause service disruption for secondary users. Balancing protection and usability requires careful timing design.

Channel availability: Secondary users need alternative channels to move to when displaced. In congested environments, finding available alternatives may be difficult.

Security: DSA systems may be vulnerable to attacks such as false sensing reports, database manipulation, or primary user emulation. Security measures add complexity and overhead.

CBRS: A DSA Case Study

The Citizens Broadband Radio Service (CBRS) in the 3.5 GHz band provides a real-world example of dynamic spectrum access:

Three-tier access: Incumbent users (primarily Navy radar) receive full protection. Priority access licensees obtain interference protection through small-area licenses sold by auction. General authorized access users may use spectrum not needed by the other tiers.

Spectrum access system: A SAS coordinates all non-incumbent use, receiving incumbent activity information and managing channel assignments to protect incumbents while maximizing availability for other users.

Environmental sensing capability: A network of sensors detects incumbent radar and reports to the SAS, enabling dynamic protection zones that adapt to actual radar activity rather than static exclusion zones.

CBRS demonstrates that dynamic sharing frameworks can work in practice, enabling commercial mobile broadband deployment while protecting incumbent government operations.

White Space Characteristics

Television bands, typically in the VHF and UHF ranges, offer favorable propagation for coverage applications:

Propagation advantage: Lower frequencies propagate farther and penetrate buildings better than higher frequency alternatives. This makes white spaces attractive for rural broadband and Internet of Things applications.

Variable availability: White space availability varies dramatically by location. Rural areas may have most TV channels available, while urban areas near multiple broadcast stations have limited white space.

Channel bandwidth: Television channels are typically 6-8 MHz wide, suitable for broadband applications without aggregating multiple narrower channels.

Co-channel restrictions: White space devices must protect television reception, requiring careful power and location management, especially near the edges of broadcast coverage areas.

Database-Driven Access

White space access is managed through geolocation databases rather than sensing alone:

Database query: Before transmitting, white space devices query an authorized database with their location. The database returns a list of available channels and authorized power levels.

Incumbent protection: Databases incorporate broadcast station coverage areas, receive site locations, and other incumbent information to calculate which channels are available at each location.

Registration: Fixed white space devices may be registered in the database, enabling other devices to avoid interfering with them.

Update requirements: Devices must requery the database periodically to account for changes in incumbent operations or database information.

The database approach avoids the sensing challenges that would arise from detecting television signals at very low levels while providing more predictable access than sensing-only systems.

White Space Applications

Several applications have emerged for white space spectrum:

Rural broadband: The long-range propagation of VHF/UHF frequencies enables cost-effective coverage of sparsely populated areas. A single tower can serve a large area that would require many towers at higher frequencies.

Smart grid communications: Utilities use white spaces for meter reading, distribution automation, and other grid management applications requiring wide-area coverage.

Internet of Things: Low-power white space devices can communicate over long distances, suitable for agricultural sensors, environmental monitoring, and asset tracking.

Campus and enterprise networks: In locations with good white space availability, these frequencies can provide backup or supplementary connectivity.

Licensed Shared Access

Licensed shared access (LSA) provides a formal framework for spectrum sharing between an incumbent and one or more additional users:

Sharing agreement: The incumbent and new entrant negotiate terms under which sharing will occur, including protection requirements, notification procedures, and any restrictions on secondary use.

LSA repository: Information about incumbent operations is maintained in a central repository. This may include geographic areas of operation, frequencies used, and time patterns.

LSA controller: New entrant base stations query the controller for available spectrum based on the repository information. The controller manages spectrum assignments and coordinates vacating when required.

LSA provides more predictable access than opportunistic systems while still enabling spectrum sharing that static allocation would not permit.

Advanced Interference Management

Rather than simply avoiding interference, advanced techniques manage interference as a resource:

Interference alignment: Through careful signal design, multiple transmitter-receiver pairs can share spectrum by aligning interference at each receiver into a reduced subspace, leaving room for the desired signal.

Coordinated multipoint (CoMP): Multiple base stations coordinate their transmissions to manage interference between cells. What would be interference from adjacent cells becomes useful signal through joint processing.

Full duplex: Advanced self-interference cancellation allows radios to transmit and receive simultaneously on the same frequency, potentially doubling spectral efficiency.

Massive MIMO: Large antenna arrays enable precise beamforming that directs energy toward intended receivers while minimizing interference to others, enabling more aggressive frequency reuse.

Coexistence Mechanisms

When different technologies share spectrum, coexistence mechanisms prevent mutual interference:

Listen-before-talk: Devices sense the channel before transmitting and defer if activity is detected. This is the basis of WiFi coexistence and is now applied to LTE in unlicensed bands (LAA).

Duty cycle limits: Regulations may limit the percentage of time devices can transmit, ensuring gaps for other users to access the spectrum.

Adaptive power control: Transmit power is adjusted based on observed interference levels, balancing coverage needs against coexistence requirements.

Frequency hopping: Spreading transmissions across multiple frequencies in a pseudo-random pattern reduces the probability of sustained interference with any single victim.

Flexible Use Licensing

Traditional service-specific licensing restricts spectrum to designated uses regardless of demand. Flexible use licensing removes technology and service restrictions:

Technology neutrality: Licenses permit any technology meeting interference requirements, enabling deployment of new systems as they become available.

Service flexibility: Licensees can provide any service, adapting to market demand rather than regulatory categories.

Secondary markets: Allowing license transfer and leasing enables spectrum to flow to those who can use it most efficiently.

Flexible licensing has been widely adopted for mobile spectrum and is expanding to other bands.

Incentive Mechanisms

Regulators are developing mechanisms to incentivize efficient spectrum use:

Use-it-or-share-it provisions: Spectrum that sits unused may be made available to others, incentivizing active use or voluntary sharing.

Spectrum fees: Fees based on bandwidth and coverage area create financial incentives to release unneeded spectrum or use it intensively.

Efficiency standards: Requirements for minimum spectral efficiency encourage adoption of advanced technologies and discourage spectrum hoarding.

Refarming incentives: Support for clearing legacy systems and migrating to more efficient technologies accelerates the transition to modern systems.

International Coordination

Spectrum efficiency benefits from international harmonization:

Equipment economies: Global frequency harmonization enables equipment to be manufactured for worldwide markets, reducing costs and accelerating technology deployment.

Roaming and interoperability: Common spectrum arrangements enable international roaming and cross-border services.

Interference management: Coordinated sharing frameworks work better when neighboring countries adopt compatible approaches.

International coordination occurs through the ITU and regional organizations, though achieving consensus among diverse interests is often challenging.

Artificial Intelligence Applications

Machine learning offers new possibilities for spectrum management:

Predictive modeling: AI can predict spectrum usage patterns, enabling proactive rather than reactive sharing decisions.

Interference classification: Machine learning systems can automatically classify detected signals, improving spectrum awareness.

Optimization: Complex spectrum sharing scenarios with many variables benefit from AI-driven optimization that exceeds human analytical capabilities.

Anomaly detection: AI can identify unusual spectrum activity that might indicate interference, unauthorized use, or security threats.

Higher Frequency Exploitation

As lower frequencies become congested, attention turns to higher bands:

Millimeter wave: Bands above 24 GHz offer vast bandwidth, though propagation challenges limit range. 5G networks increasingly use millimeter wave for high-capacity applications.

Terahertz: Research explores even higher frequencies for short-range, very high data rate applications.

New sharing models: Different propagation at higher frequencies may enable different sharing approaches than traditional bands.

Integrated Access and Backhaul

Using the same spectrum for both user access and network backhaul improves overall efficiency:

Self-backhauling: 5G systems can use the same spectrum for connecting small cells to the network core and serving users, adapting the balance based on demand.

Dynamic allocation: Spectrum can shift between access and backhaul functions based on real-time traffic patterns.

Reduced infrastructure: Eliminating the need for separate backhaul spectrum simplifies deployment and reduces costs.