Дисертації з теми "Low-bitrate"

This master thesis discusses the design and the implementation of an own developed wavelet-based codec for both video and image compression. The codec is specifically designed for low bitrate video with minimum complexity for use in online gaming environments. Results indicate that the performance of the codec in many areas equals or even surpasses that of the international JPEG 2000 standard. We believe that it is suitable for any situation where low bitrate is desirable, e.g. video conferences and mobile communications. The game development company Moosehill Productions AB has shown great interest in our codec and its possible applications. We have also implemented an existing audio solution for low bandwidth use.
Wavelet-baserad bild/video kompression.
Jonas Nilsson, Jesper Nilsson Lovägen 13, 37250 Kallinge tel: 0709708617

A large amount of the information in conversations come from non-verbal cues such as facial expressions and body gesture. These cues are lost when we don't communicate face-to-face. But face-to-face communication doesn't have to happen in person. With video communication we can at least deliver information about the facial mimic and some gestures. This thesis is about video communication over distances; communication that can be available over networks with low capacity since the bitrate needed for video communication is low. A visual image needs to have high quality and resolution to be semantically meaningful for communication. To deliver such video over networks require that the video is compressed. The standard way to compress video images, used by H.264 and MPEG-4, is to divide the image into blocks and represent each block with mathematical waveforms; usually frequency features. These mathematical waveforms are quite good at representing any kind of video since they do not resemble anything; they are just frequency features. But since they are completely arbitrary they cannot compress video enough to enable use over networks with limited capacity, such as GSM and GPRS. Another issue is that such codecs have a high complexity because of the redundancy removal with positional shift of the blocks. High complexity and bitrate means that a device has to consume a large amount of energy for encoding, decoding and transmission of such video; with energy being a very important factor for battery-driven devices. Drawbacks of standard video coding mean that it isn't possible to deliver video anywhere and anytime when it is compressed with such codecs. To resolve these issues we have developed a totally new type of video coding. Instead of using mathematical waveforms for representation we use faces to represent faces. This makes the compression much more efficient than if waveforms are used even though the faces are person-dependent. By building a model of the changes in the face, the facial mimic, this model can be used to encode the images. The model consists of representative facial images and we use a powerful mathematical tool to extract this model; namely principal component analysis (PCA). This coding has very low complexity since encoding and decoding only consist of multiplication operations. The faces are treated as single encoding entities and all operations are performed on full images; no block processing is needed. These features mean that PCA coding can deliver high quality video at very low bitrates with low complexity for encoding and decoding. With the use of asymmetrical PCA (aPCA) it is possible to use only semantically important areas for encoding while decoding full frames or a different part of the frames. We show that a codec based on PCA can compress facial video to a bitrate below 5 kbps and still provide high quality. This bitrate can be delivered on a GSM network. We also show the possibility of extending PCA coding to encoding of high definition video.

Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1996.
Includes bibliographical references (leaves 45-47).
by Rajesh Suryadevara.
M.S.

The packing and coding of digital video is a part of science where much innovation has taken place during the last few decades. The MPEG standards of video encoding are some of the most well-known and used video coding standards today. Since MPEG defines exact requirements for the decoder, but not the encoder, encoders can be made in many different ways and levels of complexity, as long as they produce legal MPEG streams that can be viewed on any MPEG-conformant decoder. This thesis will examine the possible performance of MPEG, in particular MPEG-1/MPEG-2 full TV resolution (720*576), for coding video at bitrates significantly lower than the 2-15 Mb/s MPEG-2 originally was designed for. For this purpose, encoding methods previously proposed by various researchers are presented. Furthermore a few new algorithms, which can be used for MPEG coding in general, but was constructed with a low-bitrate encoder in mind, were developed. Finally objective video quality benchmarks were conducted and the results evaluated.

This thesis introduces a coding scheme for very low bitrate video coding through the aid of principal component analysis. Principal information of the facial mimic for a person can be extracted and stored in an Eigenspace. Entire video frames of this persons face can then be compressed with the Eigenspace to only a few projection coefficients. Principal component video coding encodes entire frames at once and increased frame size does not increase the necessary bitrate for encoding, as standard coding schemes do. This enables video communication with high frame rate, spatial resolution and visual quality at very low bitrates. No standard video coding technique provides these four features at the same time.

Theoretical bounds for using principal components to encode facial video sequences are presented. Two different theoretical bounds are derived. One that describes the minimal distortion when a certain number of Eigenimages are used and one that describes the minimum distortion when a minimum number of bits are used.

We investigate how the reconstruction quality for the coding scheme is affected when the Eigenspace, mean image and coefficients are compressed to enable efficient transmission. The Eigenspace and mean image are compressed through JPEG-compression while the while the coefficients are quantized. We show that high compression ratios can be used almost without any decrease in reconstruction quality for the coding scheme.

Different ways of re-using the Eigenspace for a person extracted from one video sequence to encode other video sequences are examined. The most important factor is the positioning of the facial features in the video frames.

Through a user test we find that it is extremely important to consider secondary workloads and how users make use of video when experimental setups are designed.

Multi-view Video Coding (MVC) is vital for low bitrate applications that have constraints in bandwidth, battery capacity and memory size. Symmetric and mixed spatial-resolution coding approaches are addressed in this thesis, where Prediction Architecture (PA) is investigated using block matching statistics. Impact of camera separation is studied for symmetric coding to define a criterion for the best usage of MVC. Visual enhancement is studied for mixed spatial-resolution coding to improve visual quality for the interpolated frames by utilising the information derived from disparity compensation. In the context of symmetric coding investigations, camera separation cannot be used as a sufficient criterion to select suitable coding solution for a given video. Prediction architectures are proposed, where MVC that uses these architectures have higher coding performance than the corresponding codec that deploys a set of other prediction architectures, where the coding gain is up to 2.3 dB. An Adaptive Reference Frame Ordering (ARFO) algorithm is proposed that saves up to 6.2% in bits compared to static reference frame ordering when coding sequence that contains hard scene changes. In the case of mixed spatial-resolution coding investigations, a new PA is proposed that is able to save bitrate by 13.1 Kbps compared to the corresponding codec that uses the extended architecture based on 3D-digital multimedia. The codec that uses hierarchical B-picture PA has higher coding efficiency than the corresponding codec that employs the proposed PA, where the bitrate saving is 24.9 Kbps. The ARFO algorithm has been integrated with the proposed PA where it saves bitrates by up to 35.4 Kbps compared to corresponding codec that uses other prediction architectures. Visual enhancement algorithm is proposed and integrated within the presented PA. It provides highest quality improvement for the interpolated frames where coding gain is up to 0.9 dB compared to the corresponding frames that are coded by other prediction architectures.

I requisiti di molte applicazioni IoT necessitano di trasmettere dati su lunghe distanze, con basso data rate e con il minor impatto possibile sul consumo energetico. Le tecnologie LPWAN (Low Power Wide Area Network) sono state progettate per complementare ed in alcuni casi sostituire le soluzioni offerte dalla reti cellulari e dalle reti di sensori a corto/medio raggio. Nonostante la pletora di standards LPWAN disponibili sul mercato, la tecnologia LoRa/LoRaWAN sta riscuotendo notevole successo grazie alle performance che riesce a garantire. L’imponente mole di dati generata dalle applicazioni IoT richiede soluzioni in grado di poter archiviare e gestire in maniera efficiente il ciclo di vita delle informazioni. L’utilizzo di piattaforme di data analytics basate su sistemi NoSQL permettono una gestione più agile dei Big Data. In questa tesi è stata progettata ed implementata un’infrastruttura per il monitoraggio ambientale tramite LoRaWAN e la relativa piattaforma di data analytics adoperata per lo studio delle metriche relative alla trasmissione radio LoRa. I risultati ottenuti dalla sperimentazione possono essere usati per fare tuning delle configurazioni per il deploy in contesti reali.

Le domaine de la compression de contenus visuels (image, vidéo, éléments graphiques 2D/3D) a connu, depuis maintenant plus de vingt ans, un essor considérable avec l’émergence notamment au fil des années de nombreuses normes internationales comme JPEG, JPEG2000 pour les images fixes ou les différentes versions de standards MPEG-1/2/4 pour les données vidéo et graphiques.L’apparition des smartphones et l’explosion des applications qui leur sont dédiées a également bénéficié de ces avancées, l’image étant aujourd’hui omniprésente dans un contexte de mobilité/itinérance. Néanmoins, cela nécessite toujours des réseaux fiables et disponibles, offrant un débit suffisant pour la transmission effective de ces données visuelles qui sont intrinsèquement gourmandes en bande passante. Si aujourd’hui les pays développés bénéficient de réseaux mobiles (3G, 4G…) hautement performantes, cela n’est pas le cas d’un certain nombre de régions du monde, en particulier dans les pays émergents, où les communications s’appuient encore sur des réseaux 2G SMS. Transmettre de contenus visuels dans un tel contexte devient un défi ambitieux, qui nécessite la mise en œuvre de nouveaux algorithmes de compression. Le défi à relever consiste à assurer une transmission des images sur une bande passante correspondant à un ensemble relativement réduit (10 à 20) de SMS (140 octets par SMS).Pour répondre à ces contraintes, de multiples pistes de développement ont été envisagées. Après un état de l’art des techniques de compression traditionnelles et de leurs améliorations futures, nous avons finalement orienté nos travaux vers des méthodes de deep learning, visant à réaliser des post-traitements pour améliorer la qualité des contenus compressés.Nos contributions s’articulent autour de la création d’un nouveau schéma de compression, incluant les codecs existants ainsi qu’un panel de briques de post-traitement permettant une meilleure exploitation des contenus fortement compressés. Ces briques sont des réseaux de neurones profonds dédiés, qui réalisent des opérations de super-résolution et/ou de réduction d’artéfacts de compression, spécifiquement entraînés pour répondre aux objectifs ciblés. Ces opérations interviennent du côté du décodeur et peuvent être interprétées comme des algorithmes de reconstruction d’images à partir de versions fortement compressées. Cette approche présente l’avantage de pouvoir s’appuyer des codecs existants, particulièrement légers et peu coûteux en ressources. Dans nos travaux, nous avons retenu le format BPG, qui fait état de l’art dans le domaine, mais d’autre schémas de compression peuvent être également considérés.Concernant le type de réseaux de neurones, nos recherches nous ont conduits vers les réseaux antagonistes génératifs (Generative Adversarials Nets–GAN), qui s‘avèrent particulièrement adaptés pour des objectifs de reconstruction à partir de données incomplètes. Plus précisément, les deux architectures retenues et adaptées à nos objectifs sont les réseaux SRGAN et ESRGAN. L’impact des différents éléments et paramètres impliqués, comme notamment les facteurs de super-résolution utilisés et les fonctions de pertes, sont analysés en détails.Enfin, une dernière contribution concerne l’évaluation expérimentale. Après avoir montré les limitations des métriques objectives, qui peinent à prendre en compte la qualité visuelle de l’image, nous avons mis en place un protocole d’évaluation subjective. Les résultats obtenus en termes de scores MOS (Mean Opinion Score) démontrent pleinement la pertinence des approches de reconstruction proposées.Enfin, nous analysons une ouverture de nos travaux à des cas d’utilisation différents, d’une nature plus grand public. C’est notamment le cas pour le traitement de contenus de grande résolution plus ou moins compressés et même pour l’amélioration de la qualité de vidéos
The field of visual content compression (image, video, 2D/3D graphics elements) has known spectacular achievements for more than twenty years, with the emergence numerous international standards such as JPEG, JPEG2000 for still image compression, or MPEG-1/2/4 for video and 3D graphics content coding.The apparition of smartphones and of their related applications have also benefited from these advances, the image being today ubiquitous in a context of mobility. Nevertheless, image transmission requires reliable and available networks, since such visual data that are inherently bandwidth-intensive. While developed countries benefit today from high-performance mobile networks (3G, 4G...), this is not the case in a certain number of regions of the world, particularly in emerging countries, where communications still rely on 2G SMS networks. Transmitting visual content in such a context becomes a highly ambitious challenge, requiring the elaboration of new, for very low bitrate compression algorithm. The challenge is to ensure images transmission over a narrow bandwidth corresponding to a relatively small set (10 to 20) of SMS (140 bytes per SMS).To meet such constraints, multiple axes of development have been considered. After a state-of-the-art of traditional image compression techniques, we have oriented our research towards deep learning methods, aiming achieve post-treatments over strongly compressed data in order to improve the quality of the decoded content.Our contributions are structures around the creation of a new compression scheme, including existing codecs and a panel of post-processing bricks aiming at enhancing highly compressed content. Such bricks represent dedicated deep neural networks, which perform super-resolution and/or compression artifact reduction operations, specifically trained to meet the targeted objectives. These operations are carried out on the decoder side and can be interpreted as image reconstruction algorithms from heavily compressed versions. This approach offers the advantage of being able to rely on existing codecs, which are particularly light and resource-efficient. In our work, we have retained the BPG format, which represents the state of art in the field, but other compression schemes can also be considered.Regarding the type of neural networks, we have adopted Generative Adversarials Nets-GAN, which are particularly well-suited for objectives of reconstruction from incomplete data. Specifically, the two architectures retained and adapted to our objectives are the SRGAN and ESRGAN networks. The impact of the various elements and parameters involved, such as the super-resolution factors and the loss functions, are analyzed in detail.A final contribution concerns experimental evaluation performed. After showing the limitations of objective metrics, which fail to take into account the visual quality of the image, we have put in place a subjective evaluation protocol. The results obtained in terms of MOS (Mean Opinion Score) fully demonstrate the relevance of the proposed reconstruction approaches.Finally, we open our work to different use cases, of a more general nature. This is particularly the case for high-resolution image processing and for video compression

Ce travail de thèse répond à la problématique du transfert de la vidéo à haute résolution sur les réseaux bas débit. Dans le contexte du projet européen E3, des scénarios de téléconsultation, télésurveillance et télé-enseignement, sont définis et mis en place via des outils et des services de télémédecine. Par ailleurs, deux solutions concrètes sont proposées. Une première partie est consacrée à l’évaluation des performances de la nouvelle norme d’encodage vidéo HEVC dans le contexte médical. Les métriques objectives et les notes subjectives valident les améliorations effectuées sur cette norme par rapport au standard AVC-H.264, montrant qu’on peut gagner jusqu’à 54% en débit de compression pour une même qualité acceptable par les experts. Malgré la complexité de son architecture, une configuration adaptée au contexte bas débit (< 3 Mbits/s), a été définie et recommandée en utilisant l’encodeur temps réel x265. Une deuxième solution consiste à l’utilisation d’une méthode originale de dissimulation de données, basée sur une approche zonale, afin d’insérer des données médicales dans les séquences endoscopiques. Par rapport à l’état de l’art, cette méthode est plus performante en capacité d’insertion, imperceptibilité et complexité. Cette technique nous offre la possibilité d’insérer, en temps réel, jusqu’à 3 Mbits de données dans une vidéo médicale FHD de 10s, sans ajouter de débit ou de temps de traitement supplémentaire à l’encodeur « x265 ». La solution est améliorée par la saillance visuelle, en dissimulant dans les zones saillantes plutôt que dans toute l’image
This thesis deals with HD video transmission over low-bandwidth networks. In the context of the European project E3, scenarios such as remote consultation, telemonitoring and remote lecture, have been defined and implemented thanks to telemedicine tools and services. A first part was devoted to assess the performance of the new video encoding standard HEVC in the medical context. Objective metrics and subjective scores validated the improvements offered by this standard compared to the AVC-H.264 standard, showing that we could save up to 54% in terms of compression bitrate for a same quality, acceptable by experts. Despite the complexity of its architecture, a configuration adapted to the low bit rate context (<3 Mbps/s) was defined and recommended by using the x265 real-time encoder. A second solution has been proposed: an original method of data hiding, based on a zonal approach, to hide medical data into the endoscopic sequences. Compared to the state of the art, this method offers more efficient performance in payload, imperceptibility and complexity. This technique allows us to hide, in real time, up to 3 Mbits of data in a 10s FHD medical video, without requiring more bandwidth or an additional processing time to the encoder “x265”. The solution has been still improved by visual saliency techniques, by hiding in salient areas rather than throughout the entire image

碩士
國立中央大學
通訊工程學系
102
In recent years, lots of high resolution devices, such as smart phones, tablet, and digital camera…etc. Video compression technology becomes an important issue due to the significant increasing of data storage. A new generation of video compression standard HEVC has better performance than H.264. HEVC has improved video quality and compression efficiency, but the coding complexity also increases a lot. In this thesis, we proposed Simple Zero Block Mode Decision(S_ZBMD) in inter prediction to reduce HEVC computation complexity. We early terminate the CU size by using zero block distribution to select suitable block size and prediction mode for coding time saving. In order to improve the coding efficiency of standard HEVC, we use Modified Sum of Absolute Bi-Prediction Differences algorithm (MSABPD) to save transmitted bits. We eventually combine MSABPD and S_ZBMD to achieve both efficiency improving and time saving. Experimental results show that our proposed algorithm achieve 30.6% time saving and 2.56% BDBR saving compared with HEVC encoder in low bitrate.

High resolution surveillance video cameras are invaluable resources for effective crime prevention and forensic investigations. However, increasing communication bandwidth requirements of high definition surveillance videos are severely limiting the number of cameras that can be deployed. Higher bitrate also increases operating expenses due to higher data communication and storage costs. Hence, it is essential to develop low complexity algorithms which reduce data rate of the compressed video stream without affecting the image fidelity. In this thesis, a computer vision aided H.264 surveillance video encoder and four associated algorithms are proposed to reduce the bitrate. The proposed techniques are (I) Speeded up foreground segmentation, (II) Skip decision, (III) Reference frame selection and (IV) Face Region-of-Interest (ROI) coding. In the first part of the thesis, a modification to the adaptive Gaussian Mixture Model (GMM) based foreground segmentation algorithm is proposed to reduce computational complexity. This is achieved by replacing expensive floating point computations with low cost integer operations. To maintain accuracy, we compute periodic floating point updates for the GMM weight parameter using the value of an integer counter. Experiments show speedups in the range of 1.33 - 1.44 on standard video datasets where a large fraction of pixels are multimodal. In the second part, we propose a skip decision technique that uses a spatial sampler to sample pixels. The sampled pixels are segmented using the speeded up GMM algorithm. The storage pattern of the GMM parameters in memory is also modified to improve cache performance. Skip selection is performed using the segmentation results of the sampled pixels. In the third part, a reference frame selection algorithm is proposed to maximize the number of background Macroblocks (MB’s) (i.e. MB’s that contain background image content) in the Decoded Picture Buffer. This reduces the cost of coding uncovered background regions. Distortion over foreground pixels is measured to quantify the performance of skip decision and reference frame selection techniques. Experimental results show bit rate savings of up to 94.5% over methods proposed in literature on video surveillance data sets. The proposed techniques also provide up to 74.5% reduction in compression complexity without increasing the distortion over the foreground regions in the video sequence. In the final part of the thesis, face and shadow region detection is combined with the skip decision algorithm to perform ROI coding for pedestrian surveillance videos. Since person identification requires high quality face images, MB’s containing face image content are encoded with a low Quantization Parameter setting (i.e. high quality). Other regions of the body in the image are considered as RORI (Regions of reduced interest) and are encoded at low quality. The shadow regions are marked as Skip. Techniques that use only facial features to detect faces (e.g. Viola Jones face detector) are not robust in real world scenarios. Hence, we propose to initially detect pedestrians using deformable part models. The face region is determined using the deformed part locations. Detected pedestrians are tracked using an optical flow based tracker combined with a Kalman filter. The tracker improves the accuracy and also avoids the need to run the object detector on already detected pedestrians. Shadow and skin detector scores are computed over super pixels. Bilattice based logic inference is used to combine multiple likelihood scores and classify the super pixels as ROI, RORI or RONI. The coding mode and QP values of the MB’s are determined using the super pixel labels. The proposed techniques provide a further reduction in bitrate of up to 50.2%.

High resolution surveillance video cameras are invaluable resources for effective crime prevention and forensic investigations. However, increasing communication bandwidth requirements of high definition surveillance videos are severely limiting the number of cameras that can be deployed. Higher bitrate also increases operating expenses due to higher data communication and storage costs. Hence, it is essential to develop low complexity algorithms which reduce data rate of the compressed video stream without affecting the image fidelity. In this thesis, a computer vision aided H.264 surveillance video encoder and four associated algorithms are proposed to reduce the bitrate. The proposed techniques are (I) Speeded up foreground segmentation, (II) Skip decision, (III) Reference frame selection and (IV) Face Region-of-Interest (ROI) coding. In the first part of the thesis, a modification to the adaptive Gaussian Mixture Model (GMM) based foreground segmentation algorithm is proposed to reduce computational complexity. This is achieved by replacing expensive floating point computations with low cost integer operations. To maintain accuracy, we compute periodic floating point updates for the GMM weight parameter using the value of an integer counter. Experiments show speedups in the range of 1.33 - 1.44 on standard video datasets where a large fraction of pixels are multimodal. In the second part, we propose a skip decision technique that uses a spatial sampler to sample pixels. The sampled pixels are segmented using the speeded up GMM algorithm. The storage pattern of the GMM parameters in memory is also modified to improve cache performance. Skip selection is performed using the segmentation results of the sampled pixels. In the third part, a reference frame selection algorithm is proposed to maximize the number of background Macroblocks (MB’s) (i.e. MB’s that contain background image content) in the Decoded Picture Buffer. This reduces the cost of coding uncovered background regions. Distortion over foreground pixels is measured to quantify the performance of skip decision and reference frame selection techniques. Experimental results show bit rate savings of up to 94.5% over methods proposed in literature on video surveillance data sets. The proposed techniques also provide up to 74.5% reduction in compression complexity without increasing the distortion over the foreground regions in the video sequence. In the final part of the thesis, face and shadow region detection is combined with the skip decision algorithm to perform ROI coding for pedestrian surveillance videos. Since person identification requires high quality face images, MB’s containing face image content are encoded with a low Quantization Parameter setting (i.e. high quality). Other regions of the body in the image are considered as RORI (Regions of reduced interest) and are encoded at low quality. The shadow regions are marked as Skip. Techniques that use only facial features to detect faces (e.g. Viola Jones face detector) are not robust in real world scenarios. Hence, we propose to initially detect pedestrians using deformable part models. The face region is determined using the deformed part locations. Detected pedestrians are tracked using an optical flow based tracker combined with a Kalman filter. The tracker improves the accuracy and also avoids the need to run the object detector on already detected pedestrians. Shadow and skin detector scores are computed over super pixels. Bilattice based logic inference is used to combine multiple likelihood scores and classify the super pixels as ROI, RORI or RONI. The coding mode and QP values of the MB’s are determined using the super pixel labels. The proposed techniques provide a further reduction in bitrate of up to 50.2%.

碩士
國立臺灣大學
電機工程研究所
84
The videophone applications over communication channels is a dream to most people. Now the ITU-T Study Group XV has now drafted such a standard called H.263 suitable for video transmission below 64 kilobits per second. Thus it will be general in a few years. With today's equipment how do we implement it? There are a test model and a simulator by Telenor Research. But how do we adopt it to the resource we have ? In this thesis we change the motion estimation algorithm to a fast algorithm and enhance its performance approach the full search, but save a lot of time. We exploit the property of the input data in the quantization step to reduce dramatically the data need to be processed. We also tune some code segments such as the DCT and IDCT parts. At last we show a H.263 encoding system suitable to run in today's personal computer.

In ultra-low power Wireless Sensor Networks (WSNs) sensor nodes need to interact, depending on the application, even at a rapid pace while preserving battery life. Wireless communication brings thereby quite the burden as the radio transceiver requires a relative huge amount of power during both transmission or reception phases. In WSNs with on demand communication, the sensor nodes are required to maintain responsiveness and to act the sooner they receive a request, reducing the overall latency of the network. The aspect is more challenging in asynchronous WSN as the receiver possesses no information about the packet arrival time. In a purely on-demand communication, duty-cycling shows little to almost no improvement. The receiving node, in such scheme, is expected to last for years while also being accessible to other peers. Here arises the utility of an external ultra-low power radio receiver known as Wake-up Receiver (WuRx). Its essential task is to remain as the only part of the system running while the rest of the systems enters the lowest power mode (i.e., sleep state). Once a request signal is received, it notifies the host processor and other peripherals for an incoming communication. With the sensor node being in sleep state (WuRx active only), substantial power levels can be achieved. If the WuRx is able to interact rapidly, the added latency remains negligible. As crucial performance figures, the sensitivity and bit rate are immediately affected by the extreme low-power budget at diifferent magnitudes, depending mainly on the incorporated architecture. This thesis focuses on the design of a feature-balanced WuRx. The passive radio frequency architecture (PRF) relies on passive detection while consuming zero power to extract On-Off-Keying (OOK) modulated envelopes. The featured sensitivity, however, is reduced compared to more complex architectures. A WuRx based on PRF architecture can effectively enable short-range applications. The sensitivity can vary with respect to several parameters including the total generated noise, circuit technology and topology. Two variants of the PRF WuRxs are introduced with the baseband amplifier being the main change. The first revision employs a high performance amplifier with reduced average energy consumption, thanks to a novel power gating control. The second variant focuses on employing an ultra-low power baseband amplifier as it is expected to be in a continuous active state. This thesis also brings the necessary analysis on the passive front-end with the intention to enhance the overall WuRx sensitivity. Proof of concepts are embedded in sensor node boards and feature -61 dBm and -64 dBm of sensitivity for the first and the second variant, respectively, at a packet-error-rate (PER) of 1% whilst demanding a similar power of 7.2 µW during packet listening. During packet decoding, the first variant demands a 150 µW of power, caused greatly by the baseband amplifier. The achieved latency is less than 30 ms and the bit rate is 4 kbit/s, Manchester encoding. For long-range applications, a higher sensitivityWuRx is proposed based on Tuned-RF (TRF) architecture. By embedding a low-noise amplifier (LNA) in the receiver chain, very weak radio signal can be detected. TheWuRx emphasizes higher sensitivity of -90 dBm. The design of the LNA prioritized the highest gain and lowest bias current by sacrifcing the linearity that poses little impact on signal integrity for the OOK modulated signals. The total active power consumption of the TRF WuRx is 1.38 mW. In this work, a fast sampling approach based on power gating protocol allows a drastic reduction in energy consumption on average. By being able to sample in matter of few microseconds, the WuRx is able to detect the presence of a packet and return to sleep state right after packet decoding. Being power-gated dropped the average power consumption to 2.8 µW at a packet detection latency of 32 ms for less than 2 s of interval time between communication requests. The proposed solutions are able to decode a minimum length of 16-bit pattern and operate in the license-free ISM band 868 MHz. This thesis also includes the analysis and implementation of low-power front-end building blocks that are employed by the proposed WuRx.:1 Introduction 1.1 Motivation 1.2 Wake-up Receiver Design Requirements 1.2.1 Energy Consumption 1.2.2 Network Coverage and Robustness 1.2.3 Wake-up Packet Addressing 1.2.4 WuPt Detection Latency 1.2.5 Hosting System, Form-factor and Fabrication Technology 1.3 Thesis Organisation 2 Wireless Sensor Networks 2.1 Radio Communication 2.1.1 Electromagnetic Spectrum 2.1.2 Link Budget Analysis 2.2 Asynchronous Radio Receiver Duty-cycle Control 2.2.1 B-MAC and X-MAC Protocols 2.2.2 Energy and Latency Analysis 2.3 Power Supply Requirements 2.3.1 Low Self-discharge Battery 2.3.2 Energy Harvester 2.4 Summary 3 State-of-the-Art of Wake-up Receivers 3.1 Wake-up Receiver Architectural Analysis 3.1.1 Passive RF Detector 3.1.2 Classical Radio Architectures 3.2 Wake-up Receiver Back-end Stages 3.2.1 Baseband Amplifiers 3.2.2 Analog to Digital Conversion 3.2.3 Wake-up Packet Decoder 3.3 Power Consumption Reduction at Circuit Level 3.3.1 Power Gating 3.3.2 Interference Rejection and Filtering 3.4 Summary 4 Proposal of Novel Wake-up Receivers 4.1 Ultra-low Power On-demand Communication in Wireless Sensor Networks: Challenges and Requirements 4.2 Passive RF Wake-up Receiver 4.3 Power-gated Tuned-RF Wake-up Receiver 5 Low-power RF Front-end 5.1 Narrow-band Low-noise Amplifier (LNA) 5.1.1 Topology 5.1.2 Voltage Gain 5.1.3 Stability 5.1.4 Noise Figure 5.1.5 Linearity 5.2 Envelope Detector 5.2.1 Theory of Square-law Detection and Sensitivity Analysis 5.2.2 Single-Diode Envelope Detector 5.2.3 Voltage Multiplier Envelope Detector 5.3 Hardware Assessment 5.3.1 LNA 5.3.2 Envelope Detector 5.4 Summary 6 Passive RF Wake-up Receiver 6.1 Circuit Implementation 6.1.1 Address Decoder 6.1.2 Envelope Detector 6.1.3 Power-gated Baseband Amplifier 6.1.4 Ultra Low-power Baseband Amplifier 6.2 Experimental Results 6.2.1 Wireless Sensor Node 6.2.2 Measurements 6.3 Summary 7 Power-gated Tuned-RF Wake-up Receiver 7.1 Power-gating Protocol 7.2 Circuit Design 7.2.1 Radio Front-end 7.2.2 Data Slicer 7.2.3 Digital Baseband 7.3 Performance Evaluation 7.4 Summary 8 Conclusion 8.1 Performance Summary 8.2 Future Perspective 8.3 Applications A Two-tone Simulation Setup B Diode Models and Simulation Setup C Preamble Detection C Code Implementation Bibliography Publications
In drahtlosen Sensornetzwerken (WSNs) mit extrem geringem Stromverbrauch müssen Sensorknoten je nach Anwendung kurze Latenzzeiten erreichen ohne die Batterielebensdauer zu beeinträchtigen. Die drahtlose Kommunikation bringt dabei eine ziemliche Belastung mit sich, da der Funktransceiver sowohl während der Sende- als auch der Empfangsphase relativ viel Strom benötigt. Einige marktfähige Funktransceiver benötigen durchschnittlich ca. 10 mA im Empfangsmodus sowie 30 mA im Sendemodus. Deshalb wird heutzutage das sogenannte Duty-Cycling mit bestimmten Sende-, Empfangs- und langen Schlafzeitintervallen eingeführt. Während der Schlafphase ist der Empfänger nicht ansprechbar. Was wiederum zu einer massiven Erhöhung der Latenzzeit führen kann. In vielen Anwendungen und insbesondere im Rahmen der Digitalisierung von Prozessen wird mittlerweile die Fähigkeit On-Demand mit sehr kurzen Latenzzeiten zu kommunizieren verlangt. Diese Anforderung steht in einem Wiederspruch zum genannten Duty-cycle Betrieb. Um dieses Dilemma zu lösen wird im Rahmen dieser Doktorarbeit ein Funkempfänger mit extrem geringen Stromverbrauch untersucht und entwickelt. Mit Hilfe des extrem niedrigen Stromverbrauches kann der Funkempfänger ständig empfangsbereit sein. Er wird zum Hauptempfänger mit dem hohen Stromverbrauch zugeschaltet, so dass nur nach Aufforderung der Hauptempfänger aktiv sein wird. Dieser Empfänger wird Wake-up Empfänger (WuRx) genannt. Seine wesentliche Aufgabe besteht darin, als einziger Teil des Gesamtknotens aktiv zu sein, während der Rest in den Modus mit dem niedrigsten Stromverbrauch versetzt wird. Sobald ein Anforderungssignal empfangen wird, weckt er den Haupt-Prozessor und andere Peripheriegeräte über eine eingehende Kommunikation. Somit ist der Aufweckempfänger essenziell für die Zuverlässigkeit der drahtlosen Kommunikation. Sein Stromverbrauch sollte im µA Bereich sein. Seine Empfangsbereitschaft hängt entscheidend von seiner Empfindlichkeit sowie Bitrate ab. Eine Verbesserung der Empfindlichkeit und Erhöhung der Bitrate würden zwangsläufig zu einer Erhöhung des Stromverbrauches führen. Im Rahmen dieser Doktorarbeit werden unterschiedliche Architekturen von Aufweckempfängern untersucht und umgesetzt. Zusammenhänge zwischen Empfindlichkeit, Bitrate und Stromverbrauch wurden analysiert und mögliche Grenzen gezeigt. Ein wesentliches Augenmerk war dabei, Off-the-Shelf Komponenten zu verwenden. Im Rahmen dieser Doktorabeit wurden in Abhängigkeit von der zu erreichenden Reichweite und Häufigkeit der Kommunikation zwei wesentliche Architekturen mit geeigneten Empfindlichkeiten und extrem geringem Stromverbrauch entwickelt. Für kurze Reichweiten wurde eine passive Hochfrequenzarchitektur (PRF Architektur) basierend auf einer passiven Erkennung von OOK-modulierten (On-Off-Keying) Signalen mittels Hüllkurvenbildung entwickelt. Die erreichte Empfindlichkeit von ca. -64 dBm stellt eine wesentliche Verbesserung gegenüber dem Stand der Technik und Forschung mit einer Empfindlichkeit von ca. -52 dBm dar. Die Empfindlichkeit kann in Bezug auf verschiedene Parameter variieren, einschließlich des insgesamt erzeugten Rauschens, der Schaltungstechnologie und der Topologie. Zwei Varianten der PRF WuRxs wurden realisiert, wobei der Basisbandverstärker die Hauptänderung darstellt. Die erste Version verwendet einen Hochleistungsverstärker mit reduziertem durchschnittlichen Energieverbrauch dank einer neuartigen Leistungssteuerung. Die zweite Variante konzentriert sich auf die Verwendung eines Basisbandverstärkers mit extrem geringer Leistung, da erwartet wird, dass er sich in einem kontinuierlichen aktiven Zustand befindet. Diese Arbeit bringt auch die notwendige Analyse des passiven Front-Ends mit der Absicht, die allgemeine WuRx-Empfindlichkeit zu verbessern. Nachweise der Wirksamkeit sind in Sensorknotenmodulen eingebettet und verfügen über -61 dBm und -64 dBm Empfindlichkeit für die erste bzw. die zweite Variante bei einer Paketfehlerrate (PER) von 1 %, während beim Abhören von Paketen eine ähnliche Leistung von 7.2 µW gefordert wird. Während der Paketdecodierung erfordert die erste Variante eine Leistung von 150 µW, die stark durch den Basisbandverstärker verursacht wird. Die erreichte Latenz beträgt weniger als 30 ms und die Bitrate beträgt 4 kbit/s mit einer Manchester-Codierung. Für Anwendungen mit großer Reichweite wird ein WuRx mit höherer Empfindlichkeit vorgeschlagen. Dieser basiert auf einer TunedRF (TRF) -Architektur. Dabei werden sehr schwache Funksignale durch einen rauscharmen Verstärker (LNA) erkannt und verstärkt. Der WuRx erreicht eine bessere Empfindlichkeit von ca. –90 dBm. Dabei wurde das Augenmerk auf die höchste Verstärkung verbunden mit dem niedrigsten Vorspannungsstrom gelegt. Der LNA wird dann im nicht-linearen Bereich betrieben. Dieser Betriebsmodus beeinflusst nur im geringeren Maße die Signalintegrität der OOK-modulierten Signale. Der gesamte Leistungsverbrauch des TRF WuRx beträgt 1.38 mW. Um den Gesamtleistungsverbrauch im µW Bereich zu reduzieren, wird im Rahmen dieser Arbeit das sogenannte Power-Gating-Protokoll eingeführt. Dabei wird das Funkkanal zyklisch abgetastet. Der WuRx kann innerhalb von wenigen Mikrosekunden das Vorhandensein eines Pakets erkennen und direkt nach der Paketdecodierung in den Ruhezustand zurückkehren. Durch diesen Ansatz konnte der durchschnittliche Stromverbrauch bei einer Paketerkennungslatenz von ca. 32 ms innerhalb einer Abtastrate von 2 s auf 2.8 µW reduziert werden. Die vorgeschlagenen Lösungen können eine Mindestlänge von 16-Bit-Mustern decodieren und im lizenzfreien ISM-Band 868 MHz arbeiten.:1 Introduction 1.1 Motivation 1.2 Wake-up Receiver Design Requirements 1.2.1 Energy Consumption 1.2.2 Network Coverage and Robustness 1.2.3 Wake-up Packet Addressing 1.2.4 WuPt Detection Latency 1.2.5 Hosting System, Form-factor and Fabrication Technology 1.3 Thesis Organisation 2 Wireless Sensor Networks 2.1 Radio Communication 2.1.1 Electromagnetic Spectrum 2.1.2 Link Budget Analysis 2.2 Asynchronous Radio Receiver Duty-cycle Control 2.2.1 B-MAC and X-MAC Protocols 2.2.2 Energy and Latency Analysis 2.3 Power Supply Requirements 2.3.1 Low Self-discharge Battery 2.3.2 Energy Harvester 2.4 Summary 3 State-of-the-Art of Wake-up Receivers 3.1 Wake-up Receiver Architectural Analysis 3.1.1 Passive RF Detector 3.1.2 Classical Radio Architectures 3.2 Wake-up Receiver Back-end Stages 3.2.1 Baseband Amplifiers 3.2.2 Analog to Digital Conversion 3.2.3 Wake-up Packet Decoder 3.3 Power Consumption Reduction at Circuit Level 3.3.1 Power Gating 3.3.2 Interference Rejection and Filtering 3.4 Summary 4 Proposal of Novel Wake-up Receivers 4.1 Ultra-low Power On-demand Communication in Wireless Sensor Networks: Challenges and Requirements 4.2 Passive RF Wake-up Receiver 4.3 Power-gated Tuned-RF Wake-up Receiver 5 Low-power RF Front-end 5.1 Narrow-band Low-noise Amplifier (LNA) 5.1.1 Topology 5.1.2 Voltage Gain 5.1.3 Stability 5.1.4 Noise Figure 5.1.5 Linearity 5.2 Envelope Detector 5.2.1 Theory of Square-law Detection and Sensitivity Analysis 5.2.2 Single-Diode Envelope Detector 5.2.3 Voltage Multiplier Envelope Detector 5.3 Hardware Assessment 5.3.1 LNA 5.3.2 Envelope Detector 5.4 Summary 6 Passive RF Wake-up Receiver 6.1 Circuit Implementation 6.1.1 Address Decoder 6.1.2 Envelope Detector 6.1.3 Power-gated Baseband Amplifier 6.1.4 Ultra Low-power Baseband Amplifier 6.2 Experimental Results 6.2.1 Wireless Sensor Node 6.2.2 Measurements 6.3 Summary 7 Power-gated Tuned-RF Wake-up Receiver 7.1 Power-gating Protocol 7.2 Circuit Design 7.2.1 Radio Front-end 7.2.2 Data Slicer 7.2.3 Digital Baseband 7.3 Performance Evaluation 7.4 Summary 8 Conclusion 8.1 Performance Summary 8.2 Future Perspective 8.3 Applications A Two-tone Simulation Setup B Diode Models and Simulation Setup C Preamble Detection C Code Implementation Bibliography Publications