RAN-agnostic communications can identify intrinsic features of the unknown signal without any prior knowledge, with which incompatible RANs in the same unlicensed band could achieve better coexistence performance than today’s LBT-based coexistence. Blind modulation identification is its key building block, which blindly identifies the modulation type of an incompatible signal without any prior knowledge. Recent blind modulation identification schemes are built upon deep neural networks, which are limited to single-carrier signal recognition thus not pragmatic for identifying spectro-temporal OFDMA signals whose modulation varies with time and frequency. Therefore, this paper proposes RiSi, a semantic segmentation neural network designed to work on OFDMA’s spectrograms, that employs flattened convolutions to better identify the grid-like pattern of OFDMA’s resource blocks. We trained RiSi with a realistic OFDMA dataset including various channel impairments, and achieved the modulation identification accuracy of 86% on average over four modulation types of BPSK, QPSK, 16-QAM, 64-QAM. Then, we enhanced the generalization performance of RiSi by applying domain generalization methods while treating varying FFT size or varying CP length as different domains, showing that thus-generalized RiSi can perform reasonably well with unseen data.

Wi-Fi 7 introduces Multi-link operation (MLO) to enhance throughput and latency performance compared to previous Wi-Fi standards. MLO enables simultaneous transmission and reception through multiple links, departing from conventional single-link operations (SLO). To fully exploit MLO’s potential, it is essential to investigate Wi-Fi 7’s coexistence performance with legacy Wi-Fi devices. Existing approaches, however, have overlooked some crucial aspects of MLO, necessitating the development of a standards-compliant analytical framework to model the actual channel access mechanism of MLO. Therefore, this research aims to addresses this gap by constructing an accurate Markov chain model for MLO aligned with the backoff behaviors specified by the standard, and by deriving closed-form expressions for the throughput of each device type in the Wi-Fi 7 and legacy Wi-Fi coexistence scenario.

Millimeter-wave (mmWave) networks are conventionally considered to bear a fundamental coverage limitation, due to the directional beams and limited field-of-view (FoV) of the phased array antennas. In this paper, we explore an array of phased arrays (APA) architecture, which aggregates co-located phased arrays with complementary FoVs to approximate WiFi-like omni-directional coverage. We found that straightforwardly activating all the arrays may even hamper network performance. To fully exploit the APA's potential, we propose X-Array, which jointly selects the arrays and beams, and applies a dynamic co-phasing mechanism to ensure different arrays' signals enhance each other. X-Array also incorporates a link recovery mechanism to identify alternative arrays/beams that can efficiently recover the link from outage. We have implemented X-Array on a commodity 802.11ad APA radio. Our experiments demonstrate that X-Array can approach omni-directional coverage and maintain high performance in spite of link dynamics.

The prosperity of IEEE 802.11-based Wi-Fi networks aggravates cross-technology interference to IEEE 802.15.4-enabled ZigBee networks widely deployed to enable various Internet-of-Things applications. To make ZigBee communication reliable and robust even in a dense Wi-Fi environment, taming Wi-Fi interference in ZigBee networks especially from the perspective of physical layer is of paramount importance. In this context, this work takes aim to design a novel Wi-Fi interference-resilient ZigBee decoder called PolarScout, which separates collided ZigBee signal samples out of Wi-Fi interference to bootstrap ZigBee data decoding. Unlike several existing solutions which need clear signal preamble, tremendous signal strength difference between ZigBee and Wi-Fi, and Wi-Fi interference recognition in prior to ZigBee decoding, PolarScout aims at direct ZigBee decoding in a more generic and challenging case where Wi-Fi interference features a wide range of power levels and arises within a ZigBee packet at an arbitrary position. At the heart of PolarScout lies a subtle shell-shaping technique which harnesses a customized sample sequence to smooth the shell of corrupted signal samples. PolarScout then refers to the resulting shell to recover each contaminated ZigBee sample. Experimental results validate the superiority of PolarScout and its resilience to a wide range of Wi-Fi interference types.

Computation offloading in mobile clouds helps mobile users save energy and enhance performance via mobile-to-cloud migration of processing. Although there exist many approaches to computation offloading, they have not explicitly considered the energy-latency-pricing tradeoff from the viewpoint of mobile users' context, e.g., user tendency, the remaining battery level. This paper tries to capture the user-centric perspective via quality-of-experience (QoE), and formulates two important problems: mobile-to-cloud transmission scheduling of the offloaded task's data and offloading service class selection. Regarding transmission scheduling, we introduce a database-assisted optimal dynamic programming (DP) algorithm and then propose two suboptimal but computationally-efficient approximate DP algorithms, ADP and ADPe, based on the limited lookahead technique. Regarding service class selection, we consider multiple service classes with different computing power and service charge, and formulate an optimization problem to minimize the overall cost incurred during offloading. An extensive numerical analysis has revealed that ADP and ADPe achieve near-optimal performance incurring only 0.35 and 2.1 percent extra cost than the optimum on average, and enhances the QoE-aware cost by up to 2.38 times compared to the energy-only scheduling. In addition, our service class selection algorithm is shown to choose the best class according to user tendency and the remaining battery level.

Since License Assisted Access LTE (LAA-LTE) employs a conservative energy detection (ED) threshold for transmission to coexist with Wi-Fi in unlicensed bands, the spatial spectrum reuse of LAA-LTE can be significantly impaired. Such non-flexible thresholding has been introduced mainly due to ED's incapability of differentiating Wi-Fi frames from LTE frames. As a remedy, we design and develop an LAA-LTE transmitter based on a software-defined radio system proposing lightweight but effective Wi-Fi frame detection with which an LAA-LTE device can capture a Wi-Fi preamble by only using LTE's own time-domain samples. Experiments confirm the efficacy of the developed system.

License Assisted Access (LAA) LTE (LAA-LTE) is a new type of LTE that aggregates the licensed LTE bands with the unlicensed bands via carrier aggregation. To operate in unlicensed bands, LAA-LTE adopts the listen-before-talk policy and designs its channel access mechanism similar to WLAN's DCF. This paper considers an LAA-LTE eNB coexisting withasymmetric hiddenWi-Fi APs where the eNB can detect the APs while the APs cannot, which is caused by the asymmetric CCA thresholds. The behavior of such a network is modeled by a joint Markov chain (MC), using which steady-state probabilities, throughput, and channel access delay are derived analytically. An extensive evaluation confirms that the proposed analysis correctly models the dynamics of LAA-WLAN coexistence, and identifies important design guidelines for fair coexistence as follows. First, LAA-LTE should enable channel access priority class 4 to exploit its large contention window (CW). Second, LAA-LTE should re-design its CW doubling policy to restore the balance between LAA-LTE and WLAN in throughput and channel access delay. Third, to protect Wi-Fi, the maximum CW stage should be used more times by increasing the retry count.

Computation offloading is a promising application of mobile clouds that can save energy of mobile devices via optimal transmission scheduling of mobile-to-cloud task offloading. Existing approaches to computation offloading have addressed various aspects of the tradeoff between energy consumption and application latency, but none of them explicitly considered the dependency in optimization on the mobile user''s context, e.g., user tendency, the remaining battery level. This paper captures such a user-centric perspective in the energy-latency tradeoff via a quality-of-experience (QoE) based cost function, and formulates the problem of data offloading scheduling as dynamic programming (DP). To derive the optimal schedule, we first introduce a database-assisted optimal DP algorithm and then propose a suboptimal but computationally-efficient approximate DP (ADP) algorithm based on the limited lookahead technique. An extensive numerical analysis has revealed that the ADP algorithm achieves near-optimal performance incurring only 2.27% extra cost on average than the optimum, and enhances QoE by up to 4.46 times compared to the energy-only scheduling.

Femtocell base station (BS) is a low-power, low-price BS based on the cellular communication technology. It is expected to become a cost-effective solution for improving the communication performance of indoor users, whose traffic demands are large in general. We propose long-term parameter optimization schemes for open and hybrid femtocells, which maximize the average throughput of macrocell users by offloading the macro users' downlink traffic to femtocells. To achieve this goal, load balancing between femtocells and macrocells is needed, and hence, we jointly optimize the ratio of dedicated resources for femtocells as well as the femtocell service area in open access femtocell networks by numerical analysis. Then, we extend our algorithm to hybrid access femtocells, where some intra-femtocell resources are dedicated only for femtocell owners, while remaining resources are shared with foreign macrocell users. Our evaluation results show that the proposed parameter optimization schemes significantly enhance the performance of macrocell users because of the large offloading gain. The benefits provided to femtocell users are also adaptively maintained according to the femtocell users' requirements. The results in this paper also provide insights about the situations where femtocell deployment on dedicated channels is preferred to the cochannel deployment.

Femtocell is a cost-effective solution to both soaring wireless data traffic and coverage hole problems. Along with technical issues, the economic benefit which Wireless Service Providers (WSPs) and mobile users get from the femtocell service is a key to successful adoption of femtocell technology. In this paper, we study pricing issues for femtocell service using game theoretic approach. We propose a realistic and detailed pricing framework, and model the interaction between a monopolist WSP and users by a leader-follower game. We also model the competition between two WSPs in a duopoly market at the adoption of femtocell service by a non-cooperative simultaneous game. The competition between femtocells and repeaters is also considered using the same pricing framework. Analysis results present important factors and pricing strategies for economic success of femtocell technology.

Super Wi-Fi is a Wi-Fi like service over TV white spaces (TVWS) based on the dynamic spectrum access (DSA) technology. Although Super Wi-Fi is expected to achieve larger coverage than conventional Wi-Fi thanks to the superior propagation characteristics of TVWS, it suffers from smaller bandwidth than Wi-Fi (6–8 MHz versus 20 MHz) which degrades network capacity. Therefore, it is common belief that the two Wi-Fi technologies may target different applications such as Super Wi-Fi for coverage and Wi-Fi for speed. However, there is a lack of studies that rigorously analyzes and compares the performance of Super Wi-Fi and Wi-Fi to confirm such belief. To fill the gap, this paper performs a thorough analysis on the capacity of Super Wi-Fi under the scenario that a Super Wi-Fi access point (AP) coexists with a Wi-Fi AP. Comparing the downlink capacity of Super Wi-Fi and Wi-Fi reveals that Super Wi-Fi can outperform Wi-Fi at the outskirts of the Wi-Fi's coverage and Super Wi-Fi gets more beneficial when channel bonding is employed. In addition, the maximal coverage radius of Super Wi-Fi is derived with which Super Wi-Fi can achieve better average capacity than a network of densely-deployed Wi-Fi APs, where the maximal radius is up to 3.2 times larger than the coverage radius of Wi-Fi.

Super Wi-Fi refers to Wi-Fi-like Internet access via spectrum white spaces (WS), which is expected to enhance today's Wi-Fi thanks to the superior propagation characteristics of the WS compared to ISM/UNII bands. A Super Wi-Fi wireless service provider (WSP) dynamically leases and opportunistically utilizes a licensed band while it is temporarily unoccupied by its (licensed) primary users (PUs). The PUs' spectrum usage pattern presents time-varying spectrum availability, thus necessitating eviction of in-service customers upon return of PUs where the evicted customers are compensated with partial reimbursement of their service charge. This paper investigates the dynamics of a duopoly Super Wi-Fi market where two co-located WSPs compete for leasing better quality channels and for setting competitive service price to entice more customers. The channel quality is measured by the PUs' utilization factor (smaller the better). Since higher quality channels possess more WS incurring larger channel leasing cost, a WSP should strike a balance between channel quality and service tariff in maximizing its profit. The market competition is modeled as hierarchical noncooperative price- and quality-games, and their Nash Equilibria (NE) are derived. In addition, we investigate the impact of differentiated reimbursement rates and limited channel availability. Finally, we demonstrate the tradeoffs among leasing cost, customer arrival rate, and channel characteristics via numerical analyses.

We address the problem of rapidly discovering spectrum opportunities for seamless service provisioning in cognitive radio networks (CRNs). In particular, we focus on multichannel communications via channel-bonding with heterogeneous channel characteristics of ON/OFF patterns, sensing time, and channel capacity. Using dynamic programming (DP), we derive an optimal online sensing sequence incurring a minimal opportunity-discovery delay, and propose a suboptimal sequence that presents a near-optimal performance while incurring significantly less computational overhead than the DP algorithm. To facilitate fast opportunity discovery, we also propose a channel-management strategy that maintains a list of backup channels to be used at building the optimal sequence. A hybrid of maximum likelihood (ML) and Bayesian inference is introduced as well for flexible estimation of ON/OFF channel-usage patterns, which selectively chooses the better between the two according to the frequency of sensing and ON/OFF durations. The performance of the proposed schemes, in terms of the opportunity-discovery delay, is evaluated via in-depth simulation, and for the scenarios we considered, the proposed suboptimal sequence achieves a near-optimal performance with only an average of 0.5 percent difference from the optimal delay, and outperforms the previously proposed probabilistic scheme by up to 50.1 percent. In addition, the backup channel update scheme outperforms the no-update case by up to 49.9 percent.

We consider the problem of opportunistically scheduling low-priority tasks onto underutilized computation resources in the cloud left by high-priority tasks. To avoid conflicts with high-priority tasks, the scheduler must suspend the low-priority tasks (causing waiting), or move them to other underutilized servers (causing migration), if the high-priority tasks resume. The goal of opportunistic scheduling is to schedule the low-priority tasks onto intermittently available server resources while minimizing the combined cost of waiting and migration. Moreover, we aim to support multiple parallel low-priority tasks with synchronization constraints. Under the assumption that servers' availability to low-priority tasks can be modeled as ON/OFF Markov chains, we have shown that the optimal solution requires solving a Markov Decision Process (MDP) that has exponential complexity, and efficient solutions are known only in the case of homogeneously behaving servers. In this paper, we propose an efficient heuristic scheduling policy by formulating the problem as restless Multi-Armed Bandits (MAB) under relaxed synchronization. We prove the index ability of the problem and provide closed-form formulas to compute the indices. Our evaluation using real data center traces shows that the performance result closely matches the prediction by the Markov chain model, and the proposed index policy achieves consistently good performance under various server dynamics compared with the existing policies.

We consider the problem of scheduling low-priority tasks onto resources already assigned to high-priority tasks. Due to burstiness of the high-priority workloads, the resources can be temporarily underutilized and made available to the low-priority tasks. The increased level of utilization comes at a cost to the low-priority tasks due to intermittent resource availability. Focusing on two major costs, bandwidth cost associated with migrating tasks and latency cost associated with suspending tasks, we aim at developing online scheduling policies achieving the optimal bandwidth-latency tradeoff for parallel low-priority tasks with synchronization requirements. Under Markovian resource availability models, we formulate the problem as a Markov Decision Process (MDP) whose solution gives the optimal scheduling policy. Furthermore, we discover structures of the problem in the special case of homogeneous availability patterns that enable a simple threshold-based policy that is provably optimal. We validate the efficacy of the proposed policies by trace-driven simulations.

Cognitive radio (CR)-based Wi-Fi 2.0 hotspots are introduced as an attractive application of dynamic spectrum access (DSA), at which a wireless service provider (WSP) leases licensed channels via secondary market and offers Internet access to CR-enabled customers by opportunistically utilizing the leased spectrum. The CR users access the channels only when they are temporarily unoccupied by their legacy users, and pay a usage charge according to the WSP's pricing policy. In this paper, we study the profit maximization problem of a WSP by deriving the (sub)optimal control of admission (at CR user arrivals) and eviction (upon return of the legacy users) of CR users. We formulate the problem as a semi-Markov decision process (SMDP) with two quality-of-service (QoS) constraints on arrival-blocking and service-dropping probabilities, which is solved by the linear programming techniques. Using an extensive numerical analysis, we show that the derived policy achieves up to 22.5-44 percent more profit than simple complete-sharing algorithms in the tested scenarios. In addition, we evaluate the impact of the number of leased channels and pricing on the achieved profit, and study the tradeoffs between the two QoS constraints.

The whitespaces (WS) in the legacy spectrum provide new opportunities for the future Wi-Fi-like Internet access, often called Wi-Fi 2.0, since service quality can be greatly enhanced thanks to the better propagation characteristics of the WS than the ISM bands. In the Wi-Fi 2.0 networks, each wireless service provider (WSP) temporarily leases a licensed spectrum band from the licensees and opportunistically utilizes it during the absence of the legacy users. The WSPs in Wi-Fi 2.0 thus face unique challenges since spectrum availability of the leased channel is time-varying due to the ON/OFF spectrum usage patterns of the legacy users, which necessitates the eviction control of in-service customers at the return of legacy users. As a result, to maximize its profit, a WSP should consider both channel leasing and eviction costs to optimally determine a spectrum band to lease and a service tariff. In this paper, we consider a duopoly Wi-Fi 2.0 market where two co-located WSPs compete for the spectrum and customers. The competition between the WSPs is analyzed using game theory to derive the Nash Equilibria (NE) of the price (of the service tariffs) and the quality (of the leased channel, in terms of channel utilization) competitions. The NE existence condition and market entry barriers are also derived. Via an extensive numerical analysis, we show the tradeoffs between leasing/eviction cost, customer arrivals, and channel usage patterns by the legacy users.

In IEEE 802.22 Wireless Regional Area Networks (WRANs), each Base Station (BS) solves a complex resource allocation problem of simultaneously determining the channel to reuse, power for adaptive coverage, and Consumer Premise Equipments (CPEs) to associate with, while maximizing the total downstream capacity of CPEs. Although joint power and channel allocation is a classical problem, resource allocation in WRANs faces two unique challenges that has not yet been addressed: (1) the presence of small-scale incumbents such as wireless microphones (WMs), and (2) asymmetric interference patterns between BSs using omnidirectional antennas and CPEs using directional antennas. In this paper, we capture this asymmetry in upstream/downstream communications to propose an accurate and realistic WRAN-WM coexistence model that increases spatial reuse of TV spectrum while protecting small-scale incumbents. Based on the proposed model, we formulate the resource allocation problem as a mixed-integer nonlinear programming (MINLP) which is NP-hard. To solve the problem in real-time, we propose a suboptimal algorithm based on the Genetic Algorithm (GA), and extend the basic GA algorithm to a fully-distributed GA algorithm (dGA) that distributes computational cost over the network and achieves scalability via local cooperation between neighboring BSs. Using extensive simulation, the proposed dGA is shown to perform as good as 99.4-99.8% of the optimal solution, while reducing the computational cost significantly.

In cognitive radios, in-band spectrum sensing is essential for the protection of legacy spectrum users, enabling secondary users to vacate channels immediately upon detection of primary users. For in-band sensing, it is important to meet detectability requirements, such as the maximum allowed detection latency and the probability of misdetection and false alarm. In this paper, we propose key techniques for efficient in-band sensing. We first advocate the use of clustered sensor networks, and propose a periodic in-band sensing algorithm that optimizes sensing period and sensing time to meet the detectability requirements while minimizing sensing overhead. The scheme also determines the better of energy or feature detection incurring less sensing overhead at each SNR level, and derives the threshold aRSSthresholdon the average received signal strength of a primary signal above which energy detection is preferred to feature detection. We consider two key factors affecting aRSSthresholdnoise uncertainty and inter-CRN interference. aRSSthresholdappears to lie between -114.6 dBm and -109.9 dBm with noise uncertainty ranging from 0.5 dB to 2 dB, and between -112.9 dBm and -110.5 dBm with 1-6 interfering CRNs. We also investigate how strict the detection requirement must be for efficient reuse of idle channels without incurring unnecessary channel switches due to false detection of primaries.

Cognitive radio (CR) HotSpot is a typical example of commercialized dynamic spectrum access (DSA) in which a wireless service provider (WSP) provides CR end-users access to its local-area network. A WSP temporarily leases licensed channels from the primary licensees via spectrum auction, and subleases them to CR end-users or customers, each demanding a different amount of spectrum, according to its own advertised pricing policy. The CR customers use the subleased channels opportunistically when the channels are not being used by primary/legacy users, and hence, the amount of available spectrum resource varies with time. A WSP should, therefore, maximize its average profit by 'optimally' (in some sense) (1) controlling the admission/rejection of arriving CR end-users, and (2) determining a policy of evicting in-service opportunistic users at the channel to which primary users return. To our best knowledge, this is the first attempt to jointly optimize admission and eviction controls for the dynamic spectrum market. The WSP's profit maximization problem is formulated with a semi-Markov decision process (SMDP) and its corresponding linear programming (LP) setup. This problem is found to become nonlinear programming (NLP) subject to two quality-of-service (QoS) constraints on request-blocking and user-dropping probabilities, which can fortunately be converted to LP via some manipulation. Using an extensive numerical analysis, we discovered that the derived optimal policy achieves up to 81% more profit than a complete-sharing (CS) algorithm in the tested scenario. We also investigate tradeoffs between the two QoS constraints, and consider other important factors affecting WSP's profit maximization such as the number of leased channels, the end-user pricing, and the cost for reimbursing evicted customers.

In a cognitive radio network (CRN), in-band spectrum sensing is essential for the protection of legacy spectrum users, with which the presence of primary users (PUs) can be detected promptly, allowing secondary users (SUs) to vacate the channels immediately. For in-band sensing, it is important to meet the detectability requirements, such as the maximum allowed latency of detection (e.g., 2 seconds in IEEE 802.22) and the probability of mis-detection and false-alarm. In this paper, we propose an effcient periodic in-band sensing algorithm that optimizes sensing-frequency and sensing-time by minimizing sensing overhead while meeting the detectability requirements. The proposed scheme determines the better of energy or feature detection that incurs less sensing overhead at each SNR level, and derives the threshold aRSSthreshold on the average received signal strength (RSS) of a primary signal below which feature detection is preferred. We showed that energy detection under lognormal shadowing could still perform well at the average SNR < SNRwall [1] when collaborative sensing is used for its location diversity. Two key factors affecting detection performance are also considered: noise uncertainty and inter-CRN interference. aRSSthreshold appears to lie between -114.6 dBm and -109.9 dBm with the noise uncertainty ranging from 0.5 dB to 2 dB, and between -112.9 dBm and -110.5 dBm with 1~6 interfering CRNs.

Sensing/monitoring of spectrum-availability has been identified as a key requirement for dynamic spectrum allocation in cognitive radio networks (CRNs). An important issue associated with MAC-layer sensing in CRNs is how often to sense the availability of licensed channels and in which order to sense those channels. To resolve this issue, we address (1) how to maximize the discovery of spectrum opportunities by sensing-period adaptation and (2) how to minimize the delay in finding an available channel. Specifically, we develop a sensing-period optimization mechanism and an optimal channel-sequencing algorithm, as well as an environment- adaptive channel-usage pattern estimation method. Our simulation results demonstrate the efficacy of the proposed schemes and its significant performance improvement over nonoptimal schemes. The sensing-period optimization discovers more than 98 percent of the analytical maximum of discoverable spectrum-opportunities, regardless of the number of channels sensed. For the scenarios tested, the proposed scheme is shown to discover up to 22 percent more opportunities than nonoptimal schemes, which may become even greater with a proper choice of initial sensing periods. The idle-channel discovery delay with the optimal channel-sequencing technique ranges from 0.08 to 0.35 seconds under the tested scenarios, which is much faster than nonoptimal schemes. Moreover, our estimation method is shown to track time-varying channel-parameters accurately.

Spectrum sensing is essential to the realization of spectrum agility in cognitive radio (CR) networks. Although fundamental tradeoffs and theoretical limits associated with spectrum sensing have been studied extensively, there have been very few experimental studies focused on building a spectrum "sensor" with commercial off-the-shelf devices. We have therefore built a prototype of CR-based sensor implementation with off-the-shelf IEEE 802.11 devices. 1 In particular, we have explored issues in implementing a spectrum sensor, mostly at the MAC layer, as well as the difficulty in implanting sensing functions into industrial network interfaces. Our experimental results demonstrate the feasibility of building a spectrum sensor and the construction of an incumbent-detection mechanism with off-the-shelf devices. We have also identified technical difficulties, such as device-dependency in determining the detection threshold, in-band signal jamming between secondary devices, and adjacent channel interference due to out-of-band signal emission. This experimental experience has led us to suggest a sensor design guideline for commercial wireless interfaces which can also facilitate other CR-related research.

Static spectrum allocation prohibits radio devices from using spectral bands designated for others. As a result, some bands are under-utilized while other bands are over-populated with radio devices. To remedy this problem, the concept of spectrum agility has been considered so as to enable devices to opportunistically utilize others' spectral bands. In order to help realize this concept, we establish an analytical model to derive performance metrics, including spectrum utilization and spectrum-access blocking time in spectral-agile communication systems. We then propose three basic building blocks for spectral-agile systems, namely spectrum opportunity discovery, spectrum opportunity management, and spectrum usage coordination, and develop protocols for each blocks. These protocols are integrated with the IEEE 802.11 protocol, and simulated using ns-2 to evaluate the protocol overhead. The simulation results show that our proposed protocols can improve the throughput of an IEEE 802.11 wireless LAN by 90% for the simulated scenarios, and the improvements matched well our analytical model. These results demonstrate the great potential of using spectrum agility for improving spectral utilization in an efficient, distributed, and autonomous manner