Mobile communications demand for data grows almost two-fold every year and cellular capacity cannot catch up with this in every location and at every point in time. Not only will congestion situations appear more frequently and severely, but also every increase of data transmitted and every new infrastructure point rolled out in order to catch up with that demand is going to consume more total power. One approach towards greener communication is to reduce the actual demand. While technical solutions on Open Systems Interconnection (OSI) layers 1 trough 7 are already being exploited, layer 8, the user, is the original cause of the exploding demand. It is naive to assume that all data is equally important and valuable. User-in-the-loop (UIL) is a paradigm which can exactly hook in at this point, and it suggests to postpone, relocate, abandon or offload less urgent demand by immediate feedback and strong incentives like discounts and penalties in certain situations (in case of congestion). In particular, this means dynamic prices depending on time, location, and congestion (load) status, similar to smart grid for electricity utilities. Accounting and billing with dynamic prices, value-added services, and smart data pricing can be a challenge, and such accounting and billing are utterly incompatible with the flat-rate (all-you-can-consume) concept. However, cellular customers are involuntarily locked into contracts, but it is the current practice in the market, which is not in accordance with the free market spirit. Flat-rate plans are practically dead, but those still luring with this name, while having caps on the maximum allowance, sell fixed-volume packages under a smoother name instead. Ultimately, we need to re-establish usage-based-pricing as a fairer and more favourable way of charging. In this paper we motivate the dynamic pricing of UIL and the upcoming new need for usage-based tariffs. We compare a few status-quo tariff plans with our novel effective quantum (eBit) tariff, which is a usage-based pricing that maps multiple usage dimensions into only one metric, eBits. The eBits will be charged for with one constant unit price dollars/eBit to give the monthly bill. All Quality of Service (QoS)-, application-, and time-dependent dynamic pricing factors are integrated into the effective quantum, so that a user can easily understand his current consumption and the bill upto that point in time. Price differentiation per QoS class is easy with constant scaling factors. %This enables smart data pricing to be easily understood by the consumers. This principle can be extended to any other chargeable metric such as kWh of electricity and m^3 of water. This ultimately enables going greener and training awareness of all consumers.

Wireless cellular network planning benefits from accurate and realistic, yet relatively simple and manageable, spatial traffic models. User locations in cellular networks are often modeled as homogeneous (uniform) Poisson point processes (PPPs). However, the real user distributions are seldom purely homogeneous. Network users are usually concentrated at social attractors such as residential and office buildings, shopping malls, and bus stations. Wireless spectral efficiency depends significantly on the users' spatial heterogeneity, and thus relevant spatial traffic generators and models are important. In future (5G) networks, for which device-to-device (D2D), millimeter-wave (mmWave), and small-cell deployments in Heterogeneous Networks (HetNets), are promising technologies, it will become more important to have spatial traffic models which can represent the broad possibilities from completely homogeneous cases (e.g., a deterministic lattice) to extremely heterogeneous cases (e.g., highly clustered scenarios). In this paper, we study the spatial traffic heterogeneity of outdoor users in the denser areas of the city center of Paris, France. The building shape data is freely available from the OpenStreetMaps project. We measure the heterogeneity via a second-order statistic: the Coefficient of Variation (CoV) of two spatial metrics of the resulting point process: the Voronoi cell areas and the Delaunay cell edge lengths. The expected value of the CoV of these metrics allows us to study how the heterogeneity increases with the density of users. Moreover, we find that the statistical distribution of both these metrics is close to Weibull. Our results illustrate that the topology of the buildings in the city imposes a significant degree of heterogeneity on the spatial distribution of the wireless traffic.

Long term user rate in cellular networks is the product of spectral efficiency achieved and the resources (time/frequency slots) allocated. The former is related to the received SINR, while the latter is limited by the load of the associated cell. The max-SINR cell association strategy has been used in cellular networks from GSM to LTE. This strategy maximizes the possible achieved spectral efficiency but fails to account for the load imbalance. Recently, there have been several investigations on load-aware cell association as an approach to match the traffic demand with the traffic supply, in which a user may associate to a less loaded cell, even though it does not necessarily provide the maximum SINR. In other words, a user is associated with a cell to get more share of resources at the cost of lower spectral efficiency. This paper goes beyond that by proposing a new load balancing approach that can simultaneously increase the user received SINR and the share of allocated resources. This is achieved by the user-in-the-loop (UIL) paradigm, which encourages the user to move to a new location that maximizes the utility function considering the received SINR, cell load and the probability of moving. Numerical results show that the UIL can increase the mean user rate substantially in comparison to the max-SINR or the load-aware cell association strategy, and also results in a more balanced load across the network.

A recent approach in modeling and analysis of the supply and demand in heterogeneous wireless cellular networks has been the use of two independent Poisson point processes (PPPs) for the locations of base stations (BSs) and user equipment (UE). This popular approach has two major shortcomings. First, although the PPP model may be a fitting one for the BS locations, it is less adequate for the UE locations mainly due to the fact that the model is not adjustable (tunable) to represent the severity of the heterogeneity (nonuniformity) in the UE locations. Moreover, the independence assumption between the two PPPs does not capture the often-observed correlation between the UE and BS locations. This paper presents a novel heterogeneous spatial traffic modeling that allows statistical adjustment. Simple and nonparameterized, yet sufficiently accurate, measures for capturing the traffic characteristics in space are introduced. Only two statistical parameters related to the UE distribution, namely, the coefficient of variation (the normalized second moment), of an appropriately defined inter-UE distance measure, and correlation coefficient (the normalized cross moment) between UE and BS locations, are adjusted to control the degree of heterogeneity and the bias towards the BS locations, respectively. This model is used in heterogeneous wireless cellular networks (HetNets) to demonstrate the impact of heterogeneous and BS-correlated traffic on the network performance. This network is called HetHetNet since it has two types of heterogeneity: heterogeneity in the infrastructure (supply), and heterogeneity in the spatial traffic distribution (demand).

In this paper, we tackle the problem of multi-cell resource scheduling, where the objective is to maximize the weighted sum-rate through inter-cell interference coordination (ICIC). The blanking method is used to mitigate the inter-cell interference, where a resource is either used with a predetermined transmit power or not used at all, i.e., blanked. This problem is known to be strongly NP-hard, which means that it is not only hard to solve in polynomial time, but it is also hard to find an approximation algorithm with guaranteed optimality gap. In this work, we identify special scenarios where a polynomial-time algorithm can be constructed to solve this problem with theoretical guarantees. In particular, we define a dominant interference environment, in which for each user the received power from each interferer is significantly greater than the aggregate received power from all other weaker interferers. We show that the originally strongly NP-hard problem can be tightly relaxed to a linear programming problem in a dominant interference environment. Consequently, we propose a polynomial time distributed algorithm that is not only guaranteed to be tight in a dominant interference environment, but which also computes an upper bound on the optimality gap without additional computational complexity. The proposed scheme is based on the primal-decomposition method, where the problem is divided into a master-problem and multiple subproblems. We solve the master-problem iteratively using the projected-subgradient method. We also show that each subproblem has a special network flow structure. By exploiting this network structure, each subproblem is solved using the network-based optimization methods, which significantly reduces the complexity in comparison to the general-purpose convex or linear optimization methods. In comparison with baseline schemes, simulation results of the International Mobile Telecommunications-Advanced (IMT-Advanced) scenarios show that the proposed scheme achieves higher gains in aggregate throughput, cell-edge throughput, and outage probability.

Future generation (5G and beyond) cellular networks have to deal not only with an extreme traffic demand increase, but also an extreme level of heterogeneity in the distribution of that demand in both space and time. Traffic modeling in the time domain has been investigated well in the literature. In the space domain, however, there is a lack of statistical models for the heterogeneous User Equipment (UE) distribution beyond the classical Poisson Point Process (PPP) model. In this paper, we introduce a methodology for the generation and analysis of spatial traffic which allows statistical adjustments. Only two parameters, namely, Coefficient of Variation (CoV) and Correlation Coefficient, are adjusted to control the UE distribution heterogeneity and correlation with Base Stations (BSs). The methodology is applied to cellular networks to show the impact of heterogeneous network geometry on network performance.

Wireless cellular networks are undergoing severe changes due to the ever increasing demand of data rate. Additionally, the demand is more and more heterogeneous (imbalanced) in time and space. Sudden peaks in demand at a certain location have to be absorbed by the network. While operators traditionally over-provisioned their own separate network capacity in order to reduce the blocking and overload probabilities, this approach seems no longer economically viable. Instead, the idea of network virtualization (NV) emerged. One aspect of NV is that resources from all operators are pooled together. Shared and virtualized resources can be better distributed among all users compared to having separate subsets of users to separate subsets of resources. This holds especially if the demand is imbalanced among the operators, as shown in this paper. In this paper the stochastic Petri net (SPN) paradigm is used to provide with a compact model of NV resource pooling (RP). In contrast to the equivalent but tedious analysis of Markovian systems the SPN approach allows a quick numeric performance evaluation with tool support, thus olfering a strong modeling advantage. The scenarios analyzed here are networks of separate operators and resources, compared to one virtualized network. In a second step the scenario includes heterogeneity in demand, i.e., a load imbalance between the providers and results show much higher gains in this unbalance.

Cellular communications now and beyond 2020 faces a number of challenges. One of the trends is the ever increasing demand for data rate due to smart mobile devices with an estimated traffic growth of almost 100% per annum. Even with new cellular generation cycles every few years the same increase rate cannot be provided on the supply side. Neither anywhere nor anytime. The gap between supply and demand of wireless capacity will shorten and the conventional over-provisioning approach will not be possible anymore, especially during busy hours. The consequences are more frequent congestion situations with broken application traffic. The quality-of-experience will suffer as user expectations are high and steamed-up by advertising. An inadequate tariff system concentrating on flat-rates is also counterproductive for stability and energy-efficiency. In this paper the temporal user-in-the-loop (UIL) control approach is assumed. This user-centric model implements demand shaping by incentives in form of a dynamic usage-based tariff which adjusts based on the level of congestion in the busy hours. This is comparable to the smart grid operation principle. The novelty in this paper is the differentiated treatment for the exemplary service classes voice, video and data, for which new quantitative user response data is utilized. The control approach performance is calculated and results for stationary and dynamic scenarios are presented.

The focus of this paper is on studying the tradeoff between the sum efficiency and Jain's fairness index in general resource allocation problems. Such problems are frequently encountered in wireless communication systems with M users. Among the commonly-used methods to approach these problems is the one based on the α-fair policy. Analyzing this policy, it is shown that it does not necessarily achieve the optimal Efficiency-Jain tradeoff (EJT) except for the case of M=2 users. When the number of users M>2, it is shown that the gap between the efficiency achieved by the α-fair policy and that achieved by the optimal EJT policy for the same Jain's index can be unbounded. Finding the optimal EJT corresponds to solving a family of potentially difficult non-convex optimization problems. To alleviate this difficulty, we derive sufficient conditions which are shown to be sharp and naturally satisfied in various radio resource allocation problems. These conditions provide us with a means for identifying cases in which finding the optimal EJT and the rate vectors that achieve it can be reformulated as convex optimization problems. The new formulations are used to devise computationally-efficient resource schedulers that enable the optimal EJT to be achieved for both quasi-static and ergodic time-varying communication scenarios. Analytical findings are confirmed by numerical examples.

The cellular network users are demanding more traffic and expecting a ubiquitous high data rate. Several techniques are investigated to cope with this massive demand. All the current solutions are focusing on improving this issue from the supply side. However it is expected that a significant improvement on the system performance can be gained if the end users become a part of the system and not just consumers. The user-in-the-loop (UIL) spatial approach provides a solution to fulfil the increasing traffic demand by convincing users to move to locations with higher SINR values. The previous results of the UIL are promising and a higher cell spectral efficiency is achieved only by the involvement of the users and without any investment on the supply side. UIL is a user-centric approach that offers an incentive to influence users' behavior to participate in the system improvement process. In this paper, the previous ideal environment of UIL is extended by integrating the important effect of correlated shadowing to the analysis of UIL performance. The obtained results in this paper confirm the previous results and promise significant performance improvements. Results show that the achieved higher spectral efficiency and the required moving distance depend on the correlation parameters.

Increasing cellular traffic is the driving force for innovations in wireless communications. While voice traffic is not expected to increase much and does not require 4G systems, traffic for video and data applications is expected to grow with a rate of 100% per year. Smart mobile devices, tablets and laptop dongles will certainly make this a reality. On the other hand the supply side cannot grow with the same rate. Base stations, eNB, pico- and femtocells will bring more heterogeneity in space and new applications will bring more heterogeneity in demand over time. Designing for over-provisioning capacity has been the standard approach to stabilize traffic, but is will be harder and harder, with more congestion situations in time (busy hour) and space (crowded cell) which will break application traffic and give bad quality-of-experience of users. Furthermore, over-provisioning comes with more power consumption and higher financial expenditures for infrastructure and operating costs. The user-in-the-loop (UIL) approach offers a solution orthogonal to the traditional supply-only view. In addition to technical improvements, having a temporal demand control can alleviate the severity of busy-hour situations which formerly caused congestion and connection failures. Demand shaping is implemented by a dynamic usage-based tariff and adaptive rates depending on the load condition. The users in a cell are part of a closed control loop which reacts in cases of severe demand overload. In this paper three different service classes are controlled individually and results from analysis and simulation show the performance in stationary and dynamic scenarios. The economics of tariffs and dynamic prices and the resulting operator revenue on one side is compared to the dissatisfaction of rejected users and this gives decision indicators for the investment into new infrastructure. Overall this saves money, energy and turns situations of hard congestion into an elastic stationarity which is in the interest of both users and operators.

Traffic in mobile radio networks is expected to continue to increase by 100% per year. This imposes a big challenge for future generations (4G and 5G) of access technologies which were previously dimensioned for over-provisioning especially in the busy hours. Recent forecasts hint that this assumption and approach will not be tractable anymore. Instead the demand will exceed the supply more and more frequently causing congestion not only in short term but over longer periods of time. Any approach to engineer networks greener by better spectral efficiency only will fail to meet the global objective, if demand keeps on increasing faster than supply and efficiency. Instead, the User-in-the-loop approach allows to shape the demand where it originates either in space or time. This is achieved by incentives for users to change location or by dynamic tariffs to shift the use of congested resources out of the busy hours. This works as the smart grid of communications. While some work has already been done in this new field, the user behavior to the controlled demand shaping was based on assumptions. In this paper, the assumptions were confirmed by recent survey results which indicate that shaping user behavior works sufficiently well. Models of the user response are given and analysis and simulation results show the advantages and gains of this approach.

Resource scheduling in OFDMA cellular wireless networks is a powerful technique on the MAC layer. Utilizing adaptive modulation and coding allows the effective use of all signal-to-interference ratio (SINR) ranges. Typical single antenna spectral efficiency values for LTE-Advanced range between 4.8 near the base station and 0.2 b/s/Hz at the cell edge. With best-effort traffic and full buffer assumption the tradeoff between emphasizing the cell center or cell edge can be explored extensively. However, with real-time traffic present, this takes priority without fairness adjustment alternatives. In this paper the mixed traffic scenario is studied with an abstract stochastic Petri net model. The exploration of the degrees of freedom by studying the real-time traffic proportion and a fairness adjustment parameter provides new insight to the potential feasible region. The results show that the tradeoff between emphasizing the cell edge performance and maintaining a high average spectral efficiency is most powerful in the best-effort case, while an increasing level of real-time traffic reduces the room for a tradeoff. The stochastic Petri net analysis approach allows numeric analysis without simulation by utilizing Markov chain equivalence and steady state calculations. This model is deliberately abstract but flexible enough to study the tradeoff.

Cellular wireless networks based on OFDMA utilize adaptive modulation and coding to operate effectively in regions of high and low SINR. Therefore the local single antenna spectral efficiency ranges between 5 in the cell center and 0.2 b/s/Hz at the cell edge. The scheduling goal of high average spectral efficiency contradicts the goal of a good rate fairness among all terminals. Also there is a tradeoff between increasing the cell edge performance and maintaining a high average spectral efficiency. In this paper a stochastic Petri net analysis approach is taken and a numeric analysis is performed based on Markov chain equivalence and steady state calculations. The proposed models are deliberately abstract but offer commonly used tuning parameters in order to study the tradeoff without too many degrees of freedom.

In this paper, we propose a novel distributed inter-cell interference coordination (ICIC) scheme. The proposed scheme, which runs in polynomial time, finds a near-optimum dynamic resource partitioning that maximizes a proportional-fairness criterion in the entire network. The proposed scheme is based on primal-decomposition method, where the problem is divided into a master and multiple sub-problems. The master-problem is solved using projected-subgradient method while each of the sub-problems is solved using minimum-cost network flow optimization. Through extensive simulations of four IMT-advanced scenarios, we quantify the gains achieved using the proposed scheme. We demonstrate that the proposed scheme achieves both high cell-edge throughput, that is comparable to frequency reuse 3, and high aggregate throughput, that is at least as good as the aggregate throughput achieved by frequency reuse 1.

In this paper the system analysis of modern wireless systems is simplified by providing simple yet powerful models for the wireless channel in the environment of higher layer abstract system descriptions with generalized stochastic Petri nets (SPN). This modeling approach is capable of deriving performance metrics in terms of packet delays even under heterogeneous, asymmetric, bursty and underutilized traffic conditions, because they are easy to model with SPN. The missing link in wireless systems are suitable channel models, which can now be used as a plug-in submodel inside a larger composite Petri net model. A number of models are proposed, starting from the finite-state Markov channel model approach. Performance results for a multihop relayed transmission under varied traffic load show the utility of this modeling approach.

Multihop and multipoint transmissions are two of the main features towards an increased spectral efficiency for the LTE-Advanced mobile radio system. Fixed wireless relays with in-band backhauling are considered as a multihop technique in LTE-Advanced. Relays improve cell capacity and cell edge user performance depending on the deployment. In this paper, at first, results for the peak spectral efficiency of LTE-Advanced are presented and secondly, an analytical model to calculate the cell spectral efficiency of relay enhanced cell deployments in the context of the IMT-Advanced evaluation is presented. The developed model is applied to the LTE-Advanced system comparing different relay deployments with different frequency reuse schemes.

Mobile radio networks of the IMT-Advanced systems family promise ubiquitous broadband access and high area coverage, with rates of several 100 MBit/s. They claim to guarantee QoS support in terms of low delay and guaranteed throughput. However, with the availability of flat rate plans and bandwidth-hungry applications of future mobile devices the systems are facing a hard challenge in satisfying all demands at the same time in a traffic load situation which can best be characterized as total overload or full buffer. In this situation the basic voice service must still be operational to full Erlang capacity, no matter what load the data traffic offers. In this paper the hierarchical static priority scheduling scheme is used to accomplish the required separation. Candidate technologies like LTE-Advanced, WiMAX are based on OFDMA access which allows flexible radio resource allocation, but has an inherent near-far heterogeneity which leads to unfairness if it is optimized for spectral efficiency. Multi-hop relaying is one way to diminish the near-far problem, which is why it is an important part of the system concepts. The scheduling scheme is designed with this multihop capability in mind and results show that in heterogeneous scenarios a proportional fair substrategy achieves the desired fairness. To achieve QoS distinction a flow management must be provided in OSI layer two, so that flows of different classes can be treated separately. This paper treats the flow management concept for multi-hop mobile radio systems, a key enabling technology for QoS aware resource scheduling. It features the cross-layer signaling of QoS requirements by the Application Layer to the Data Link Layer.

Multihop cellular networks offer increased coverage and capacity compared to singlehop cells of the same size. This is not as huge as using more base stations (BS) instead of wireless relay nodes (RN), but by this way there is no need for a fixed network access to each node. Therefore a cost-efficient deployment is easily achieved (less CAPEX and OPEX). Even when the coverage of a BS is already quite good, relays help to increase the overall system capacity. This paper analyzes numerically with analytic background the multihop operation in the frequency division duplex mode and discusses the MAC-layer protocol performance for the 3GPP- LTE system. OFDMA with centrally controlled resource coordination is assumed here. By following the analytic model we show that for relaying, OFDMA resources are utilized effectively in time and frequency and provide good coverage and capacity results. Parameters like directional relay antennas, space or time division multiple access and relay antenna gain are studied.

The future cellular radio networks like 3G-LTE are based on an OFDMA physical layer. The duplex scheme is preferably frequency division (FDD) because of its advantages in long range. Huge area coverage in a cost-efficient way is important problem for the early deployment. Time after time the demand in densely populated areas will grow, so a higher cell capacity over the area is needed. The requirements are expensive to solve with a traditional cellular architecture, because a fibre line access will be needed at any base station location, which are sometimes only a few 100 meters apart. This paper deals with multihop operation as an option to improve the coverage as well as the capacity issue at low cost. Homogeneous relays act like base stations, but without the need of a cable or fibre access. They are simply fed by the same radio technology in their first hop. The contribution of this paper is especially the multihop operation in the FDD mode and performance results for the throughput on the MAC-layer. The results were obtained with an analytic model, numerically evaluated with Matlab.

We analyze the impacts of multi-hop relaying on the quality of service of voice transmission in the downlink OFDMA communication system. It is shown that the system performance in terms of capacity and quality strongly depends on deployment and operating parameters such as the location of relays, traffic loads and the allowable packet error rate. The impact of relaying is less evident when the traffic load is light and the base station can effectively cover the whole cell. In the high traffic load region, multi-hop relaying is most effective in reducing the outage probability while increasing the cell throughput.

Many signal processing systems use event driven mechanisms-typically based on finite state machines (FSMs)-to control the operation of computationally intensive (data flow) parts. The state machines in turn are often fueled by external inputs as well as by feedback from the signal processing portions of the system. Packet-based transmission systems are a good example for such a close interaction between data and control flow. For an efficient design flow it is of crucial importance to be able to model and analyze the complete functionality of the system within one single design environment. Therefore, we developed a computational model that integrates the specification of control and data flow by combining the notion of data flow graphs with event driven process activation.

Multirate Dataflow Graphs (MR-DFGs) are used for modelling iterative computations, allowing concurrency and arbitrary data rates at ports. This model is often used for signal processing algorithms. For static scheduling the iteration period bound represents the final barrier for the computation speed, the approximation of which is often the goal of an implementation. For the singlerate case (SR-DFG), where all rates are one, an explicit bound exists and is subject of many published papers. This work presents a bound for the multirate case, which reduces to the known bound if applied to an SR-DFG. Assumptions made are a vectorized execution and a blocked schedule that organizes multiple iterations inside one period (also called execution cycle). The influence of characteristic properties in the multirate case is emphasized and related to terms from the Petri-Nets theory.

The retiming of DSP algorithms exhibiting multirate behavior is treated. Using the non-ordinary marked graph model and the reachability theory, we provide a new condition for valid retiming of multirate graphs. We show that for a graph with n nodes the reachability condition can be split into the reachability condition for the topologically-equivalent unit-rate graph (all rates set to one), and (n/sup 2/-n)/2 rate-dependent conditions. Using this property a class of equivalent graphs of reduced complexity is introduced which are equivalent in terms of retiming. Additionally, the circuit-based necessary condition for valid retiming of multirate graphs is extended for the sufficient part.