Traffic Offloading in Unlicensed Spectrum for 5G Cellular Network: A Two-Layer Game Approach

Deployment of LTE in unlicensed bands is being considered in Rel-13 of LTE. This feature is called Licensed-Assisted Access (LAA) using LTE. Unlicensed band is attractive due to the large amount of available spectrum. However, in shared spectrum the coexistence between LAA and WiFi systems becomes a primary challenge. This paper presents an analytical framework to investigate the downlink coexistence performance between two systems with a simple listen-before-talk (LBT) mechanism enforced on LAA. Using this framework, theoretical models based on Markov chains are established for both systems and downlink throughput can be calculated. Numerical results from the models show that the simple listen-before-talk scheme is very effective in LAA and LAA coexistence scenario (i.e. two LAA systems sharing the same spectrum). In LAA and WiFi coexistence scenario, it can improve WiFi performance substantially.

In order to meet the requirement of dramatically growing mobile data, the operators improve their network by aggregating multiple carriers to increase the system capacity. Due to the limited availability of licensed spectrum, unlicensed spectrum with large amount of resources has attracted more and more world-wide interests recently. The Licensed Assisted Access (LAA) using LTE(-A) on unlicensed bands is investigated and being standardized in 3rd Generation Partner Project (3GPP) Release-13. This paper focuses on studying the co-existence techniques of multiple operators within LAA system. Although Listen-Before-Talk (LBT) mechanism is identified as a necessary feature in LAA to cope with the interference between Wi-Fi and LTE, it may not be efficient enough to mitigate the interference within LAA while between different operators. In this paper, we propose an enhanced LBT mechanism with cross-correlated information by utilizing assistant measurement report from users. System level simulation is conducted to illustrate the performance gain of the proposed method. It is shown that the enhanced LBT mechanism has outstanding performance in medium and high traffic scenarios in terms of user experienced throughput.

To meet the dramatically increasing traffic demand, licensed-assisted access (LAA) has been proposed for long-term evolution (LTE) systems to operate on the unlicensed spectrum. However, it is a huge challenge to achieve fair and efficient coexistence between LAA and Wi-Fi on the same band. In this paper, we devise a novel channel access approach for LAA eNBs to improve system throughput and achieve fair coexistence with Wi-Fi nodes. Specifically, Carrier Sense Multiple Access with Enhanced Collision Avoidance (CSMA/ECA), which is able to reach and maintain collision-free operation by deterministic backoff (DB) after successful transmissions, is introduced to the listen-before-talk (LBT) procedure of LAA eNBs to improve the coexistence system throughput. However, note that the fixed DB value in traditional CSMA/ECA cannot ensure fair coexistence all the time corresponding to different network sizes. Therefore, an adjustable DB value scheme is devised to replace the fixed one, in order to achieve fair coexistence whatever the number of Wi-Fi nodes and LAA eNBs is. Aiming to obtain the maximum system throughput improvement, the behavior of a LAA eNB is modeled as a Markov chain, and the throughput of the coexistence network is derived. Then, we formulate a throughput optimization problem and develop an algorithm of DB value adjustment, which can not only achieve the maximum throughput improvement but also maintain the fair coexistence whatever the number of nodes in the coexistence network is. At last, numerical results are presented which validate the accuracy of our analysis model and demonstrate the effectiveness of the proposed algorithm.

Long Term Evolution (LTE) in unlicensed spectrum, Licensed Assisted Access (LAA), has been considered as an effective complement in offloading growing traffic. The LAA is expected to be deployed initially through the small cell integrated under LTE macro cell. Therefore, LTE users under the coverage of LAA small cell must decide to connect one of the co-located networks either LTE or LAA to maximize their performance. In this scenario, we model a network selection game for LTE users using mixed strategy game theoretic approach. Afterwards, we formulate behavior of strategic users toward network selection as a function of the number of nearby users. The simulation results show that increasing number of nearby users make a user less likely to switch to LAA small cell. Additionally, more users nearby make it less likely that LAA resources is being utilized.

Many wireless local area network (WLAN) performance estimations are done with the assumption of uniformly distributed stations (STAs). In practice, on the contrary STAs are distributed unevenly among access points (APs), causing hot-spots and under utilized APs in a wireless network. Considering a WLAN is made up of multiple APs, having some APs carrying excessive loads (i.e. hot-spots) degrades both the considered APs as well as the overall network performance. The system performance can be improved by associating incoming STAs effectively throughout the network, in a sense to balance the network load evenly between APs and relieve the hot-spot congestion. Currently employed user association method in IEEE 802.11 WLANs considers only the received signal strength of APs at STAs, and associates STAs to the closest (in signal strength sense) AP, ignoring its load and interference value.Novel user association algorithms are required for congestion relief and network performance improvement. In this work, a new distributed association algorithm taking into consideration not only the received signal strength of the APs at STAs but also AP loadings and interference is proposed. A new AP load calculation method acknowledging the interference between STAs and APs is presented. Our simulations demonstrate that the proposed algorithm can improve the overall system throughput performance more than 50% and offers a better load distribution across the network compared to conventional association algorithm.

3rd generation partnership project (3GPP) has developed 5-GHz unlicensed band long term evolution (LTE), referred to as licensed-assisted access (LAA). LAA adopts listen-before-talk (LBT) operation, resembling Wi-Fi's carrier sense multiple access with collision avoidance (CSMA/CA), to enable collision avoidance capability. However, the frame structure overhead of each LAA downlink burst varies with the ending time of each preceding LBT operation. Given that, the existing analytic models of Wi-Fi cannot be used to evaluate the performance of LAA due to the fact that, unlike Wi-Fi, LAA frame structure overhead varies with the ending time of the preceding LBT operation. In this letter, we propose a novel Markov chain-based analytic model to cope with the variation of the LAA frame structure overhead. Comparison between analysis and simulation results shows that the difference between them is merely 0.2% on average, demonstrating the accuracy of the proposed model.

Licensed-Assisted Access (LAA) using Long Term Evolution (LTE) is a promising technique to extend the available bandwidth by aggregating the unlicensed spectrum. To guarantee fair co-existence of nodes deployed by different operators and different RATs, listen-before-talk (LBT) mechanism and carrier selection can be adopted by LAA as efficient candidates. In this paper, we provide insight into the LBT based LAA networks deployed by multiple operators using a hard core point process (HCPP) model. Moreover, an adaptive carrier selection scheme is developed where neighboring small cells select different carriers to avoid blocking each other so that larger number of active small cells can be retained after LBT channel sensing. The analytical results are developed in terms of active probability and coverage probability for single carrier, random carrier selection and the adaptive carrier selection scheme. It is revealed by numerical results that the number of active small cells can be significantly increased by employing adaptive carrier selection especially for large number of carriers, which further improves the network throughput.

Licensed-Assisted Access (LAA) which allows LTE to be deployed in unlicensed spectrum is a promising technique currently under standardization in 3GPP Rel-13. To guarantee a peaceful coexistence of LAA with the incumbent systems such as WiFi, listen-before-talk (LBT) which requires the node to sense the medium before data transmission is introduced. On the other hand, LBT also prevents simultaneous transmission from neighbouring nodes of the same LAA operator. In order to increase the spectral efficiency of LAA, frequency reuse for transmission by neighbouring nodes of the same operator has been agreed to be one target of LAA design. In this paper we evaluate the performance of LAA in an indoor scenario assuming both one and two LAA networks deployed by different operators. It is shown that LBT is indeed needed to ensure good performance with multiple LAA operators, especially at high load. And frequency reuse between neighbouring nodes of the same LAA operator can further improve the performance.

Long Term Evolution (LTE)-License Assisted Access (LAA), which leverages unlicensed resource sharing with the Wi-Fi network, is a promising technique to address the spectrum scarcity issue in present and future wireless networks. However, unlicensed spectrum sharing between Wi-Fi and LAA requires fair resource allocation with specific performance guarantees for both sets of Wi-Fi and LAA stations. In this paper, an optimal communication policy is devised for LAA stations coexisting on a single unlicensed channel with Wi-Fi stations. The inter-network collisions are avoided through non-overlapping transmission phases for Wi-Fi and LAA networks. The throughput performance of LAA network is maximized while guaranteeing a proportionally fair performance among LAA stations and a fair share for Wi-Fi stations. The proposed scheme, unlike the state-of-the-art coexisting mechanism, jointly optimizes the transmission probability and the transmission rate for each LAA station. The formulated optimization problem to maximize network throughput is solved analytically. The numerical results demonstrate a significant improvement in the LAA throughput, more than 75%, as compared to the case when transmission probabilities are not optimized. Moreover, a notable gain of 8-9% in the fairness index reflects the intra-network fairness of the proposed LAA network over the conventional LAA network.

Coexistence with wireless-fidelity (Wi-Fi) is a key aspect for licensed assisted access (LAA) to succeed. The main coexistence mechanism between both technologies is listen before talk (LBT), whereby radio frequency energy is sensed over a short period of time and compared to an energy detection (ED) threshold. If the sensed energy is below the ED threshold, a node is then allowed to transmit. Otherwise, it backs off. Given the default Wi-Fi and LAA ED thresholds, and the lack of virtual carrier sensing at the LAA side, coexistence problems may raise at below such ED thresholds. In order to address this issue, we propose a practical and easy implementable adaptive ED threshold algorithm, which can dynamically tune the ED threshold at the Wi-Fi and LAA nodes. Simulations results in an enterprise scenario show that the Wi-Fi performance can be significantly enhanced, while moderately impacting LAA performance.

Due to the urgent demands for wider bandwidth to support continuously blooming wireless applications, deploying Long-Term Evolution Advanced (LTE-A) networks to unlicensed spectrum has been regarded as one of the most crucial remedies to tackle the shortage of available bandwidth on licensed spectrum. This craving consequently drives the development of 3GPP Licensed-Assisted Access (LAA) to sustain both downlink and uplink transmissions of LTE-A (known as LAA networks) on 5 GHz unlicensed spectrum. With vast bandwidth to accommodate a considerable number of bands, simultaneously utilizing multiple bands (carriers) on unlicensed spectrum is of practical interests to enhance data rates, and has been supported in downlink transmissions of LAA. However, to maintain fairness among coexisting users in sharing unlicensed radio resources, transmissions on unlicensed spectrum are restricted by regional regulations such as listen-before-talk (LBT). Particular procedures are also framed by LAA to impose barriers of channel access if an eNB wishes to utilize more than one carriers. Due to the lack of an analytical model to characterize the performance of exploiting multiple carriers, it is a critical challenge for an eNB to determine the optimum number of utilized carriers, which has been the primary concern obstructing the practice of multi-carrier access. In this paper, we consequently provide an analytical model capturing the performance in terms of the number of successful channel accesses (throughput) for an eNB utilizing multiple carriers, to serve the major demand of enhancing data rates. Taking co-channel interference from WiFi and other eNBs into account, the provided model facilitates an eNB to estimate the throughout under given WiFi activity, number of deployed eNBs, and number of utilized carriers. The provided performance analysis thus paves essential foundations for further performance optimization in practical LAA deployment.

Small cell deployment in heterogeneous networks, whereby small cell base stations (SBS) are deployed alongside traditional macro-cell base stations, is a proven solution for enhancing spatial frequency reuse across licensed spectrum in long-term evolution (LTE) networks. In order to mitigate the shortage of licensed spectrum resources, licensed-assisted access (LAA) has been introduced to allow LTE SBSs to share the unlicensed channel with WiFi nodes. As such, a complex yet interesting optimization problem results from the joint utilization of licensed and unlicensed spectrum resources by the SBSs to meet the quality-of-service (QoS) requirements of small cell users (SUEs). In this paper, we highlight the fundamental tradeoff induced by the SBSs between the amount of co-channel interference (CI) resulting from the reuse of licensed spectrum resources and the collision probability (CP) imposed on the co-existing WiFi nodes due to the sharing of unlicensed spectrum resources in such a coexisting LTE LAA-WiFi heterogeneous network deployment. We find that this fundamental tradeoff can be analyzed by developing a power allocation rule with double water-filling lines and the complete set of Pareto optimal solution can be achieved by the weighted Tchebycheff method. Our simulation results show that the proposed joint resource allocation algorithm can achieve a flexible and suitable tradeoff between the licensed spectrum CI and the WiFi CP according to the QoS requirements of SUEs in LTE LAA networks.

The fifth-generation (5G) mobile network is facing severe challenges of the traffic demand surge and spectrum scarcity. To improve the network capacity and data rate, the deployment of Long Term Evolution (LTE) cellular system on unlicensed band is considered as one of the promising solutions. But the interference problem among LTE and WiFi system on the same unlicensed band hinders the deployment process, which is a critical issue unsolved for network operators. Therefore, a novel channel access scheme integrating the energy detection scheme and back-off mechanism in LTE system has been proposed in this paper, to minimize the interference to WiFi system. The throughput of both LTE and WiFi systems and the optimal sensing time and back-off time are theoretically analyzed with proofs. Furthermore, an efficient adaptive reconfiguration strategy of access parameters is proposed based on the channel occupancy cognition results. Simulation results verify that the proposed channel access scheme with optimal back-off time can double the throughput of WiFi system when coexisting with LTE system on unlicensed band.

Based on LBT (Listen Before Talk) in 4G LAA (License Assisted Access), this paper proposes a scheme to detect and avoid broadband interference in power wireless private network. Without modifying the existing technical specifications of power wireless private network, such as LTE-G or IoT-G, it can effectively detect and avoid broadband interference through scheduling to improve the signal quality of power wireless private network using the proposed algorithm in BS (Base Station) side.

In order to coexist with the WiFi systems in the unlicensed spectrum, Long Term Evolution (LTE) networks can utilize periodically configured transmission gaps. In this paper, considering a time division duplex (TDD)-LTE system, we propose a Q-Learning based dynamic duty cycle selection technique for configuring LTE transmission gaps, so that a satisfactory throughput is maintained both for LTE and WiFi systems. By explicitly taking the impact of IEEE 802.11n beacon transmission mechanism into account, we evaluate the coexistence performance of WiFi and LTE using the proposed technique. Simulation results show that the proposed approach can enhance the overall capacity performance by 19% and WiFi capacity performance by 77%, hence enabling effective coexistence of LTE and WiFi systems in the unlicensed band.