請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/66356
標題: | 基於FPGA的部分可自我重組生醫訊號處理系統 FPGA-based Partial Self-Reconfigurable Bio-Signal Processing System |
作者: | Sih-Jie Liu 劉思杰 |
指導教授: | 張璞曾(Fok-Ching Chong) |
關鍵字: | 部分重置,部分可自動重置,感測器網路,心律,哮鳴音,現場可程式邏輯閘陣列, partial reconfiguration,partial self-reconfiguration,wireless sensor network,body sensor network,WSNs,BSNs,ECG,heart rate,wheeze,FPGA, |
出版年 : | 2012 |
學位: | 碩士 |
摘要: | 無線個人感測器網路(Wireless Body Sensor Networks)應用於生理監測上通常需要在身上裝置多組感測模組,此類系統通常講求可攜性(Portable)、可穿戴性 (Wearable)、低功耗(Low Power Consumption)以及能夠同時即時監測多種不同生理訊號。
以此需求,本研究提出以現場可規劃邏輯閘陣列(Field Programmable Gate Array, FPGA)為基礎的生醫感測平台,此平台可以自行與多種無線生醫裝置連線,感應接收不同種類的生理訊號,如數位資料、醫電訊號、體音訊號等,並進行訊號處理或資料壓縮。此平台在運作的過程中,如果感應到新的裝置,可使用運行中部分重置技術(Run-Time Partial Reconfiguration, RTPR) 在不影響系統正常運作的前提之下對邏輯區塊進行重置,以符合處理新進生理訊號的需求,而在處理生理訊號的過程中如有遇到緊急狀態,亦可更新FPGA中的邏輯區塊為緊急模式。與傳統相同性質的微處理器嵌入式系統相較而言,使用數位積體電路的優勢在於其高速運算能力,能夠解決更為複雜、高資料量的運算需求,且不受微處理器運算能力之限制;尤其在同時處理多個任務的情形之下,使用積體電路能夠達到平行處理,擁有傳統嵌入式系統所無法比擬的優勢,而兼具可重置與積體電路兩大特性的FPGA無疑是最好的選擇。 在此論文中,我們挑選常見的醫電訊號─心電圖及運算量大的體音訊號─肺音來實作,以證明此系統的可靠度和即時訊號處理。我們將心律偵測演算法以及哮鳴音偵測演算法以硬體實現,達到即時的數位訊號處理,同時利用RTPR技術,以最小的硬體成本支出來模擬多位受測者、多種生理訊號同時運行於有限資源下之情境;與傳統相同設計的數位電路相較之下,本系統可節省超過10% 的硬體資源消耗,同時省電模式的設計能夠讓系統在閒置狀態下的消耗功率顯著降低,此外可程式邏輯電路的特性亦使得設計者在設計以及維護上具有極大的彈性及可塑性,證明RTPR技術適用於硬體成本、體積及消耗功率受限的各種生醫感測系統領域。 Wireless Body Sensor Networks (WBSNs) applied on physiological signal monitoring usually consist of several sensor modules that are pasted on human body. These kind of systems are usually requested to have portable, wearable, and low power consumption features, and can real-time monitor several bio-signals simultaneously. In this thesis, we proposed a bio-signal sensing platform based on Field Programmable Gate Array (FPGA) to fit above requests. The platform can automatically connect to several wireless bio-medical devices to receive various physiological signals, such as digital data, medical electronic signal, sound within body, and etc., and then act signal processing or data compression. If the platform during runtime senses a new bio-medical device, it can utilize the Run-Time Partial Reconfiguration (RTPR) technology of FPGA to dynamically replace the logic blocks during runtim, thus it can change a new algorithm of signal processing for a new coming physiological signal. Furthermore, if the physiological signal during processing is recognized as in emergency situation, the platform will immediately update logic blocks as emergency mode. Compared with traditional microprocessor based embedded system which has similar funtion, the advantage of digital integrated circuits is their high speed computing ability. It let the digital integrated circuit able to solve the needs that more complex and larger amount of data, and not limitation of the computing ability of microprocessor. However, use of the integrated circuit has incomparable advantage with traditional microprocessor based embedded system – the exploitation of parallel working tasks, espically while the system needs to handle multiple tasks simultaneously. And the FPGA has both characters: reconfigurable design and integrated circuit properly be the best choice. In this research, we chose popular medical electronic signal - ECG and high operational sound within body - loung sound to implement. We realized the heart rate and wheeze detection algorithms on FPGA to achieve real-time digital signal processing. And utilizing RTPR to realize running multi-users’ multiple bio-signal processing programs simultaneously in the same resource-limited device that uses least of the hardware resource. Compared with traditional system which has the same digitial circuit design, this system can save more than 10% hardware resource, and the desgin of power saving mode can significant reduce the power consumption. Besides, the feature of programmable logic circuit also provides great flexibility and plasticity in desgin and maintenance aspects. These prove the RTPR technology suit for many bio-signal processing systems whose hardware space, cost and power consumption are limited. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/66356 |
全文授權: | 有償授權 |
顯示於系所單位: | 生醫電子與資訊學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-101-1.pdf 目前未授權公開取用 | 5.12 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。