Human activity recognition (HAR) with WiFi channel state information (CSI) is attractive for privacy-preserving, device-free sensing, yet real deployments still struggle with three coupled issues: robustness across rooms and bands, efficiency on edge hardware, and unified support for multiple tasks. We present UN-2DCNN, a lightweight 2D-CNN pipeline tailored to indoor, multi-user CSI sensing. The design reduces temporal redundancy via a simple temporal skipping augmentation, learns a compact 128-D representation with a small CNN+GAP backbone, and injects reliability feedback through uncertainty-aware feature scaling (UAFS): Stage-1 predictive entropy is mapped to a gating weight that rescales features before a second decision head. A channel-attention MLP further suppresses spurious subcarrier responses. Evaluated on a recent multi-user CSI benchmark across classrooms, meeting rooms, and empty environments at 2.4/5 GHz, UN-2DCNN consistently outperforms competitive RNN/Transformer baselines while using only ∼1M parameters and maintaining sub-2 s test-time latency. Beyond higher accuracy, the model exhibits faster, smoother convergence and improved calibration (fewer overconfident errors). Ablations confirm that removing attention, UAFS, or the second-stage head yields consistent drops, and simple temporal skipping on the data side complements model-side selectivity. These results indicate that reliability-aware, lightweight designs can deliver practical accuracy–efficiency trade-offs for CSI-based perception on edge/IoT platforms.

Wi-Fi Channel State Information (CSI)-based human activity recognition (HAR) leverages rich channel propagation characteristics to enable non-intrusive, privacy-preserving, and device-free sensing. In multi-user wireless environments, however, HAR faces significant challenges, including multi-path interference, signal overlap, label ambiguity, cross-domain channel variability, and constraints imposed by real-time deployment on resource-limited edge devices. This paper presents MuECNet (Multi-user Enhanced Convolutional Network), a lightweight and modular deep learning framework designed to operate under realistic multi-user, multi-activity CSI sensing conditions. MuECNet integrates three key components: (i) an Enhanced Convolutional Encoding Module (ECEM) for fine-grained temporal–spectral feature extraction that preserves channel propagation signatures; (ii) a Branch-wise Feature Normalization (BFN) module for user-specific channel representation learning; and (iii) an adaptive Decision Module for multi-label activity inference. To improve robustness under diverse and dynamic channel conditions, we introduce MixUp-based data augmentation to emulate activity overlap and reduce label ambiguity. Evaluations on the WiMANS dataset show that MuECNet achieves 64.46% (2.4 GHz) and 64.15% (5 GHz) accuracy with only 1.19M parameters, 1.74G FLOPs, and 0.28 s inference latency, outperforming baseline models such as ABLSTM and THAT while reducing model size by up to 75%. Ablation studies confirm the contribution of each module, with accuracy drops of up to 6.41% when removed. These results demonstrate that MuECNet provides a robust and communication-efficient solution for integrating CSI-based sensing into future wireless networks and IoT systems.

Low-power wide area networks (LPWANs) are wireless communication systems that enable long-distance communication with low power consumption, and have attracted attention as a wireless infrastructure for Internet of Things (IoT) systems. However, the long-distance nature of LPWANs can cause the hidden node problem, which results in a lower packet delivery ratio (PDR). To address the hidden node problem, this paper proposes a channel assignment method. First, we derive an approximate formula for PDR under the impact of hidden nodes, and then propose an optimal channel assignment method based on simulated annealing, using the derived formula. Simulation results confirm that the proposed method can mitigate the degradation of PDR and increase the capacity of LPWAN networks.

Wi-Fi-based Human Activity Recognition (HAR) using Channel State Information (CSI) has received significant attention due to its non-intrusive nature and wide applicability in smart cities, healthcare, and smart home systems. Despite its potential, real-world deployment faces a major challenge: ensuring the model’s ability to generalize across different users, who exhibit distinct behaviors and encounter diverse environmental factors. This issue, termed Cross-User Generalization (CUG), arises from distribution shifts caused by variations in users’ body shapes, postures, movements, and environmental conditions. To address this challenge, we propose a cross-domain generalization framework based on an Attention-based Bidirectional Long Short-Term Memory (ABLSTM) network. The proposed framework enhances the generalization capabilities of HAR models by leveraging multi-source training and incorporating attention mechanisms for effective temporal feature learning. Specifically, the model learns user-invariant features across diverse training users and adapts to new users without requiring direct exposure to their data. Extensive experiments on a multi-user CSI dataset demonstrate that the ABLSTM model improves average test accuracy by 2.0%–6.7% compared to BiLSTM and GRU models, and over 18.0% compared to MLP-based models, while achieving the lowest training and testing loss. This methodology offers a robust and scalable solution for Wi-Fi-based HAR, promoting reliable deployment in complex, multi-user environments encountered in smart cities, healthcare monitoring, and security surveillance.

Low-power wide-area networks (LPWANs), which achieve low-power consumption, enabling long-term battery operation and long-range communication capabilities, have emerged as a new standard for realizing massive wireless sensor networks (WSNs). LPWANs are becoming increasingly popular due to low introduction costs, which stem from features, such as using unlicensed bands and low-cost nodes. LPWANs are particularly useful for Internet of Things (IoT) applications that periodically collect information about specific observation targets. However, LPWAN generally adopts a simple medium access control (MAC), which significantly degrades communication quality due to packet collisions when the traffic load increases. Thus, MAC design is critical for realizing large-scale LPWANs. Carrier sense multiple access (CSMA) can autonomously avoid packet collisions. However, its performance is drastically deteriorated due to the hidden node problem in large-scale LPWANs. This article proposes an autonomous distributed MAC strategy that can suppress the hidden node problem by utilizing traffic periodicity. The proposed method is designed carefully considering LPWAN-specific constraints, such as duty cycle limitations in unlicensed bands, low clock accuracy of nodes, and limited downlink communication opportunities. From numerical results, the proposed method improves the packet delivery rate (PDR) performance by up to approximately by 29%, 9%, and 8% compared to ALOHA, CSMA-x, and the state-of-the-art LoRa MAC, respectively.

The evolution of sensor networks is indispensable in modern society, with environmental sensing technologies playing a critical role. Packet communication, commonly used in many systems, is not suited for efficient transmission of environmental information due to the excessive data volume of headers. This paper focuses on the physical wireless parameter conversion sensor networks (PhyC-SN), examining how this innovative approach addresses the issues of packet communication and significantly reduces energy consumption. PhyC-SN allows for simultaneous environmental recognition through the frequency distribution of received signals by switching the carrier frequencies according to sensor information. However, PhyC-SN cannot generate heatmaps that depict the spatial distribution of sensor information. Thus, this study proposes a new method that combines Mobile Data Gathering Stations (MDGS) with PhyC-SN to enable the generation of heatmaps. By leveraging dynamic position control of MDGS and aggregation of sensor data, this combination allows for precise environmental recognition linked with sensor location information. This integration facilitates the construction of an energy-efficient sensor network, offering a sustainable technology solution.

Low Power Wide Area Networks (LPWANs), including Long-Range Wide Area Network (LoRaWAN), that enable long-distance communication with low-power consumption have been attracting attention along with the development of the Internet-of-Things (IoT). Carrier Sensing (CS) is mandated before packet transmission in some regions and countries. However, the hidden node problem becomes more serious as the communication area becomes larger. Packet-Level Index Modulation (PLIM) assigns indexes to the frequency channel and time slot of packet transmission to increase throughput. This paper proposes a PLIM-based method to enable the GateWay (GW) to obtain the sensing relationships among nodes without explicit signaling. The obtained information can be used for node clustering and resource allocation to alleviate the hidden terminal problem. Computer simulation results show that the proposed method improves Packet Delivery Rate (PDR) by up to 13% compared to the conventional method using PLIM. And can achieve a similar PDR performance as the situation where the inter-node distance is ideally known.

Low-power wide area networks (LPWANs), such as long-range wide area networks, are increasingly adopted as a communication standard for wireless sensor networks. LPWAN has been adopted for systems that periodically collect data from sensors, such as in environmental monitoring. However, continuous packet collisions may occur in periodic traffic systems owing to the adoption of a simple random-access scheme in LPWAN. In addition, the use of low-cost nodes results in clock drift, rendering the synchronization of nodes in the system challenging. Thus, this study proposes a wireless resource allocation scheme to avoid continuous packet collisions under the adverse effect of such clock drift. In the proposed scheme, the gateway determines the transmission offset and frequency channel for each node utilizing the periodic traffic feature and clock drift. Based on computer simulation results, the proposed scheme can improve the packet delivery rate by over 20% compared with benchmark methods.

In recent years, long-range wide-area network (LoRaWAN) has attracted considerable attention due to its ability to realize massive machine-type communication (MTC); however, its throughput is limited by the duty cycle (DC). Packet-level index modulation (PLIM) can increase throughput by utilizing a data packet’s frequency channel and transmission timing as the information-bearing index. In PLIM, a node selects a specific transmission resource (frequency channel and timing) within a frame and transmits a packet. If the node cannot transmit the packet at the specific transmission resource, it discards the packet. This packet discard results in throughput degradation of each node. Thus, this paper proposes an index mapping scheme that divides a frame into multiple subframes to provide each node with multiple transmission opportunities. A node performs channel activity detection (CAD) at the selected transmission resource within a subframe; if the node cannot transmit a packet, it moves to the next subframe. The proposed scheme adaptively maps the information bit sequence onto a transmission resource within each subframe based on the wireless environment and communication quality. Computer simulation results show that the proposed scheme improves throughput by increasing the transmission opportunities and reducing the packet discard rate under the constraint of DC.

With the fast growth of Internet-of-things (IoT) and machine-to-machine (M2M) communications technology, low- power wide area networks (LPWAN) such as long-range wide area network (LoRaWAN) are attracting attention. One of the main applications of LPWAN is information collected from a large number of sensor nodes, where network traffic is generally dominated by periodic uplink (UL) traffic. In LPWAN, each sensor node transmits packets at arbitrary timing, which causes packet collisions and degrades the system communication quality. Especially in the case of periodic UL, continuous packet collisions may occur. Therefore, this paper proposes a centralized radio resource allocation scheme that avoids packet collisions of periodic traffic in LPWANs, taking into account the characteristics of the periodical traffic. The computer simulation results show that the proposed scheme can improve the average packet delivery ratio (PDR) by 18% and the age-of-information (AoI) performance compared to the ALOHA protocol.

Low-power wide-area network (LPWAN), which includes the long-range wide-area network (LoRaWAN) protocol, is an enabling technology that satisfy low power consumption and long-range communication. In many applications of LPWAN, an end node (EN) transmits data at regular intervals. Based on this, we had previously proposed the concept of packet-level index modulation (PLIM) to increase the number of information bits transmitted by one data packet. In PLIM, the generation interval between two consecutive packets is split into multiple time slots. Each EN transmits its packet in a specific combination of time slot and frequency channel, which represents the index, to convey additional information bits. To retrieve additional information, the gateway (GW) detects the time slot in which the data packet is being transmitted. Therefore, accurate synchronization between EN and GW is essential. However, clock drift occurs due to the inexpensive real-time clock oscillator on each EN, which results in timing misalignment between each node and the GW. This article proposes a simple clock drift estimation and compensation method for the PLIM. An experimental measurement is performed to model the clock drift. The numerical results obtained using the clock drift model show that it can accurately detect the time slot index under the influence of clock drift. Furthermore, PLIM is implemented on a commercial LoRaWAN node and GW to demonstrate its practicability.

In low power wide area network (LPWAN), intra-system and inter-system interferences have become crucial issues due to a huge number of terminals and the use of the same frequency channels among LPWAN systems. Thus, we propose a channel allocation method for one of the LPWAN systems, long range wide area network (LoRaWAN), according to the radio environment to suppress these interferences. First, a LoRa gateway (GW) estimates an empirical cumulative distribution function (ECD F) of inter-system interference signal power in each frequency channel. Based on the obtained ECDF, the GW allocates LoRa nodes to the frequency channels in ascending order of the inter-system interference time ratio. Then, the GW reallocates the LoRa nodes to the frequency channels by a search algorithm based on a formula of intra-system interference probability so that this probability on each channel becomes lower than a threshold while maintaining the intersystem interference suppression. The simulation results show that our proposed method can suppress packet losses compared to the conventional random channel selection.

Long range wide area network (LoRaWAN) enables low-power and long-range communication suitable for wireless sensor networks (WSNs). Since LoRaWAN is a decentralized network, packet collision may occur if multiple nodes send data packets simultaneously. It is necessary to adjust the node's transmission interval properly to avoid such packet collisions. This paper proposes a transmission interval control strategy that takes advantage of an acknowledgment (ACK) signal from a gateway (GW) to each node. Instead of transmitting a controlling signal, the GW selects one of two receiving windows for ACK packet transmission to indicate 0 or 1. Thus, it does not incur any overhead. The proposed method adaptively changes the transmission interval of each LoRaWAN node by utilizing the ACK signal returned from the GW. The proposed method uses a simple prediction method that is computationally simple and can provide good properties even when there are constraints on the systems associated with observing gentle changes, such as environmental monitoring. Computer simulation results show that the proposed method can reduce the number of transmitted packets while achieving the same observation accuracy level as conventional methods.

A low power wide area network (LPWAN) is one of the promising technologies intended for Internet of Things (IoT). The LPWAN features a wide communication area that is a few kilometers to a few tens of kilometers; low power consumption is achieved using a battery. This study measures the transmission characteristics of a LoRaWAN which is one of the LPWAN. Therefore this paper shows the measured results in field experiment of certain cases: a road along the Tamagawa river in Tokyo, inside and around campuses, and a medium city environment. These measured results indicate that it is difficult to accurately perform a carrier sensing for LoRaWAN. Subsequently, we propose a method to realize the accurate spectrum sensing for spectrum sharing considering sensing sensitivity.

In the physical wireless parameter conversion sensor network (PhyC-SN), the sensor controls the frequency of carrier according to the sensing result and all the nodes inform the sensing results to the fusion center, simultaneously. The dynamic range of sensing result which node sends to FC depends on the available channels of frequency spectrum. In spectrum sharing to the other system (primary system), PhyC-SN (secondary system) considers the assignment between the quantization level of digital sensing result and the channel of frequency spectrum. In this paper, for constructing this assignment, the fusion center collects two types of tendencies. First one is the sensing results. The mean and variance value of sensing results can be estimated by data tracking technique. Second one is a channel occupancy rate (COR). From COR, the fusion center predicts the probability of finding non-occupied channel. The fusion center can construct the assignment map based on minimum mean square error between the predicted sensing result and the exact one. As a result, the constructed assignment achieves the high spectrum usage efficiency.

Recently, some papers that apply a multi-armed bandit algorithm for channel selection in a cognitive radio system have been reported. In those papers, channel selection based on Upper Confidence Bound (UCB) algorithm has been proposed. However, in those selection, secondary users are not allowed to transmit data over same channels at the same time. Moreover, they do not take security of wireless communication into account. In this paper, we propose secure channel selection methods based on UCB algorithm, taking secrecy capacity into account. In our model, secondary users can share same channel by using transmit time control or transmit power control. Our proposed methods lead to be secure against an eavesdropper compared to conventional channel selections based on only estimated channel availability. By computer simulation, we evaluate average system secrecy capacity. As a result, we show that our proposed channel selections improve average system secrecy capacity compared to conventional channel selection.

Dynamic spectrum access of cognitive radio exploits the low active channel and thus achieves the high frequency spectrum usage efficiency. Since it has freedom for access channel, the access channel of a node is different from that of the other one. It is an access channel mismatch. For compensating it, the exchanging of control signal between two nodes is required but the large delay and the low spectrum efficiency are serious problem. This paper proposes novel control signal and novel receiver for compensating access channel mismatch. In the receiver, we use nonlinear amplifier and thus the wide spread spectrum of distorted control signal can be detected even under the access channel mismatch. We use an interleaved frequency division multiple access (IFDMA) modulation scheme as control signal. IFDMA modulation has robustness to nonlinear distortion. Therefore, the peculiar feature of IFDMA modulation is used as identifying the control signal from detecting the spectrum of distorted IFDMA signal. The computer simulation shows that the highly accurate detection of control signal is achieved under the access channel mismatch and it is possible to inform the access channel number to the other node.

In this demonstration, we introduce a demonstration system of highly efficient environment aware wireless sensor network by considering characteristics of sensing information and spectrum efficiency. In this network, a suitable sensor system is selected from multiple wireless sensor systems by considering the demand of each sensor and surrounding wireless environment. In our demonstration system, we prepare 2.4GHz ZigBee, 920MHz small power wireless system and 460MHz wireless sensor network converting the sensing information to frequency called physical wireless parameter conversion sensor network (PHY-C SN). In order to understand the surrounding wireless environment, we also prepare a radio environment database based on the measurement data of each sensor. In CCNC 2016, we plan to introduce a video of the field test with more than 20 sensor nodes and to exhibit a small demonstration system based on multiple wireless sensor devices and USRPs.

A cognitive radio system coexisting with various systems must perform a spectrum sensing for finding a white space (WS). In this paper, we propose a new two-stage spectrum sensing method using an energy detection (ED) based on fast Fourier transform (FFT). In the case that an unknown signal is present the FFT band, false signals are detected in the surrounding FFT bins of the received signal due to a spectral leakage. Whereas a windowed FFT is well known as the solution of the spectral leakage, the spectrum spread to out of the original signal frequency is detected. The shape of the spectrum depends on the kinds of the window function. Therefore we improve the detection performance of ED by employing the property of the windowing in the second stage of the proposed spectrum sensing method. The cognitive node can obtain the chance of the wireless communication because WS can be detected with low false alarm probability by the proposed method.

Physical conversion sensor networks (PhyC-SN) achieve simultaneous data collection in wireless sensor networks. Although the PhyC-SN can recognize the median and the outliers of all the sensing data simultaneously, their separation into each sensor data is a difficult task. To address this task, we have proposed a data separation method that utilizes multi-target tracking with use of the sensing data and the received spectrum power for the separation. Even if some of the sensor data are close to each other, the instantaneous power of the sensor data bearing signal is a useful feature for the separation. In this paper, we clarify the separation accuracy of our separation method under various wireless environments by computer simulation.

In Physical Layer Network Coding (PLNC), since the two sources access the relay station at a time, the relay station can hardly demodulate each information owing to the mutual interference between two transmitted signals. The PLNC achieves the high secure relay communication based on physical layer security (PLS). The source needs the transmit power control based on the channel state information (CSI) for maintaining the secure mutual interference. The untrusted relay, however, may impersonate the CSI for breaking it. So far, the authors have evaluated the amount of the leaked information to the untrusted relay by the impersonation of CSI but the leaked information is not maximal. For the countermeasure to the impersonation of CSI, the most powerful model of the untrusted relay to maximizing the leaked information is necessary. This paper constructs the optimal impersonation of CSI by a linear programing problem.

Spectrum leasing is one of the cognitive radio models, where a primary user leases its spectrum resources for secondary user in exchange for cooperation. In [1] cooperative communication with beamforming (BF) in spectrum leasing was proposed for primary secure communications. However, interference is caused to the primary user from secondary user with its scheme. Therefore, we focus on cooperative communications where secondary users use orthogonal BFs that do not affect primary user. We consider the effects of secondary cooperation and the knowledge of eavesdropper's channel state information (CSI) on the cooperative communications. By simulation, we evaluate primary secrecy capacity, primary secrecy capacity outage probability, secondary sum throughput, and primary throughput. We show that secondary transmit BFs in spectrum leasing are effective for both primary and secondary systems in terms of primary secure communications without any interference from secondary user and with small degradation of the secondary throughput.

We proposed the PAPR reduction technique of resource block assignment. In the proposed technique, the effect of PAPR reduction can be controlled by RB assignment. If the PAPR reduction is more powerful, the output signal power from power amplifier is enlarged. In multi-cell environment, the cell coverage is decided by transmit power. Therefore, the access between mobile station and base station can be controlled by transmit power. When base stations cooperate each other, the base station of the mobile station's access can be switched by channel state information. As a result, the site diversity gain can be obtained. This paper considers two base station cooperation system with the PAPR reduction of RB assignment.

In this paper, we propose an improved soft-output M-algorithm (ISOMA) and use it to reduce the computational complexity of differential encoded LDPC (DE-LDPC) coded systems with multiple-symbol differential detection (MSDD). The proposed ISOMA combines the features of iterative tree search detection based on M-algorithm (ITS-MA) and soft-output M-algorithm (SOMA) approaches, which can guarantee that the log-likelihood ratio (LLR) of each coded bit can be computed with high reliability. Simulation results show that the computational complexity of the MSDD soft-input soft-output demodulator (SISOD) as well as the iterative decoding complexity of DE-LDPC coded systems with MSDD can be greatly reduced by the proposed ISOMA. Compared with the existing ITS-MA and SOMA, the proposed ISOMA has better performance in terms of the BER performance and the ability of reducing the decoding complexity of DE-LDPC systems with MSDD. In addition, it is shown that the approach of the ISOMA using a proper selected scaling factor (SF) in the computation of the extrinsic LLRs can be used to further improve the performance of the ISOMA.

Switching estimated receiver position (SwERP) scheme is proved to be a promising solution for indoor positioning system, owing to its high achievable accuracy and consistency. The high positioning accuracy is achieved from sensitivity (Rx,S) limit, field-of-view (FOV) limit, and assisted azimuth and tilt information from 6-axis sensor. In this paper, we propose a physical simulation model that can predict the latter three factors. The conventional visible light communication (VLC) model uses only geometric optics (GO) to predict light propagation, which is not enough to define the FOV limit. Thus, we propose a novel method using rotation matrix with cone function and support vector machines (SVMs) to classify the boundary of FOV limit. Based on sensitivity and FOV limits, possible azimuth and tilt angulations are mathematically defined. Moreover, by including FOV limit into the simulation, transmitters and their mirrors that are outside FOV limit can be neglected, of which reduce at least 80% of computation during GO calculation.

Nearest transmitter information is a mandatory requirement for utilizing Switching Estimated Receiver Position (SwERP) scheme, of which was proposed to enhance accuracy of Visible Light ID (VLID) positioning system. The conventional approach is achieved by controlling Sensitivity (RxS) limit. Conversely, general distance estimation by optical received power can also be used. Both of the mentioned methods have disadvantage on implementation complexity and inaccuracy due to noise fluctuation, respectively. In this paper, we propose a nearest transmitter classification (NTC) method by utilizing Optical Orthogonal Code (OOC) with On-Off Keying (OOK) modulation at transmitters (Txs), and perform oversampling at a receiver (Rx) to overcome the limitation of bandwidth resolution (B). The results confirm that the proposed method can classify the nearest Txaccurately, with some tradeoff with computation complexity from the increment of oversampling factor (Oc).

TCP service constructs high quality end-to-end communication link. TCP needs two way communication links. First link and second one are necessary for transmitting data packet from source to destination and ACK packet form destination to source, respectively. Therefore, TCP throughput strictly depends on both links. In wireless communication, the link speed is time selective due to shadowing effect. When the link speed is dramatically changed during the event of downloading, the impact of time selective link speed should be significant. In this paper, we clarify the impact of fluctuating link speed to TCP throughput by network simulator. In addition, we propose the adaptive link speed control of ACK link for achieving large ACK throughput. We also confirm the effect of proposed technique by network simulator.

In cognitive radio, the mobile terminal senses the wireless accessing of the other wireless communications, specifies the vacant frequency channel, and then tries to use the specified frequency spectrum. In sensing technique, the detection accuracy depends on the detection time and thus the number of detectable systems also depends on it. When the other system is in the undetectable area for the cognitive radio, cognitive radio must protect the wireless access of it. In this paper, we consider the cognitive radio with using carrier sensing and transmitting power control. In considered system, the accessing protocol based on carrier sensing is established for protecting detectable other system. In addition, the transmitting power for suppressing the interference to the undetectable system is controlled. Since the sensing time depends on the detectable area, the transmitting power of cognitive radio transmitter and the opportunity of finding vacant spectrum are changed. Therefore, we derive the optimal sensing time for achieving the largest channel capacity.

The novel frequency rotation techniques for suppressing the path interference and increasing the frequency diversity gain are proposed in uplink IFDMA system. In IFDMA, the path interference in multipath fading channel and the reduction of frequency diversity gain degrades the error rate performance, especially more significantly as the number of simultaneous access users becomes large. For suppressing the path interference and enlarging the frequency diversity gain, the frequency rotation has been proposed for down link access but it cannot be applied to the uplink due to the distortion of user frequency orthogonality. In the proposed frequency rotation, multiple users do not share the multiple frequency orthogonal channels and thus the orthogonality among the users can be maintained. The effect of the revised technique is shown by computer simulation.

Interleaved frequency division multiplexing (IFDM) can achieve high diversity gain as well as can establish orthogonal frequency multiplexing by means of comb-shaped frequency spectrum. In IFDM, as the number of multiplexing users becomes large, the comb-shaped frequency spectrum should be narrow, so that frequency diversity gain is decreased. In addition, IFDM suffers from inter-path interference introduced in the transmitted signal by multipath fading channel. In this paper, the frequency spectrum rotation is proposed. In the proposed method, the comb-shaped frequency spectrum is frequency-shifted for every modulated symbol. As a result, the frequency spectrum of the frame composed of many modulated symbols is widely spread. In addition, the inter-path interference can be suppressed because the modulated symbol is orthogonal in frequency with its time-consecutive symbol. From the computer simulation, the frequency rotation can achieve the better error rate performance thanks to increasing frequency diversity gain and suppressing inter-path interference.

Interleaved frequency division multiple access (IFDM A) is frequency orthogonal multiple access scheme and acquires high frequency diversity gain. In addition, the peak to average power ratio (PAPR) of IFDM symbol is small due to the single carrier modulation and thus the nonlinear amplifier having good power efficiency can be used in the mobile station. If an IFDMA is applied to the down link access (from base station to mobile station), the power consumption can be saved. However, the interleaved frequency division multiplexing (IFDM) has not been clarified yet for down link access. In this paper, we evaluate the error rate performance and PAPR performance with using the computer simulation. We derive the advantages and the disadvantages of IFDM compared to multicarrier code division multiplexing (MC-CDM).

This paper proposes an IP packet loss compensation scheme with bicast and forwarding for hard handover (HO) between different base stations (BSs). The proposed scheme could reduce the control delay for HO since it previously transfers IP packet from an access router to both an old BS and a new BS before HO, i.e., cell change is performed. Furthermore, the proposed scheme could realize loss less of IP packet by forwarding the IP packet only when IP packets transferred before starting bicast are buffered in the old BS for the HO request timing. From our computer simulation results, the proposed scheme could reduce the HO control delay at 95% cumulative distribution function (CDF) by approximately 1.7 sec, 340 msec, and 170 msec compared with the conventional forwarding scheme and by approximately 300 msec, 320 msec, and 330 msec compared with the conventional bicast scheme when the maximum Doppler frequency (fdmax) is 5.55 Hz and the data rate on a wired propagation channel is 10, 50, and 100 Mbps, respectively. Therefore, we clarified that the proposed scheme is effective as an IP packet loss compensation scheme for hard HO

In the compressing and expanding (companding) method, at the transmitter, baseband signals are compressed for reducing peak-to-average power (PAPR) and at the receiver, the received signals are expanded for removing the distortion caused by the compressing. However, since the signals are distorted by the band pass filter (BPF), the high power amplifier (HPF), and the channel fading, the distortion caused by the compressing cannot be removed by expanding the received signals. We propose the novel companding system that compensates the distortions caused by not only the compressing but also the BPF, the HPF, and the fading. From the computer simulation, we show that the proposed companding system achieves better BER than the conventional one with the same level of the spectrum in the out-of-bandwidth.

Multi-carrier code division multiple access (MC-CDMA) has been one of the candidates for the next generation wireless communication systems. In an uplink, the MC-CDMA system suffers from the different access timing, the different fading, and the different frequency offset of each active user. In this paper we analyze the effects of the frequency offset compensation with MMSE-MUD (minimum mean square error based multi-user detection) for MC-CDMA in a quasi-synchronous uplink. We consider the MC-CDMA system with two subcarrier mapping schemes, the continuous mapping scheme and the discrete mapping scheme. From our theoretical analysis and computer simulation, we show that the MMSE-MUD can compensate the different frequency offsets among users. We also show that the MMSE-MUD significantly improves the bit error rate (BER) for the MC-CDMA system with the continuous mapping scheme.