No abstract

The obligation for comprehensive fetal heart rate investigation had driven to improve the passive and non-invasive diagnostic instruments despite the USG or CTG method. Fetal phonocardiography (f-PCG) utilizing the auscultation method met the above criteria, but its interpretation frequently disturbed by the presence of noise. For instance, maternal heart and body organ sounds, fetal movements noise, and ambient noise from the environment where it is recording are the noise that corrupted the f-PCG signal. In this work, the use of discrete wavelet transforms (DWT) to eliminate noise in the f-PCG signal with SNR as the performance parameters observed. It was observing the effect of changes in wavelet type and threshold type on the SNR value. The test was carried out on f-PCG data taken from physio.net. Initial SNR values ranged from -26.7 dB to -4.4 dB; after application of DWT procedure to f-PCG, SNR increased significantly. Based on the test results obtained, wavelet type coif1 with the soft threshold gave the best result with 11.69 dB in SNR value. The coif1 had a superior result than other mother wavelets that use in this work, so the fPCG signal analysis for fetal heart rate investigation suggested to use it.

ABSTRAKAspek utama yang membedakan sensor satu dengan yang lainnya adalah tingkat akurasinya. Pada penelitian ini, dibuat sistem untuk menurunkan tingkat deviasi untuk meminimalisir kesalahan hasil pengukuran pada sensor berbiaya murah. Sensor yang digunakan adalah sensor tekanan udara BMP180. Sensor tersebut digunakan untuk mengukur kedalaman berdasarkan tekanan udara dalam air. Moving Average Filter (MAF) digunakan untuk membuang pencilan data, sehingga didapatkan data yang lebih relevan yang kemudian digunakan untuk melakukan curve fitting. Kemudian dilakukan analisis regresi linear untuk menghasilkan persamaan yang berfungsi sebagai pengoreksi data terekam dari sensor tersebut. Pengujian sistem dilakukan melalui beberapa skenario lalu diambil persamaan yang menghasilkan nilai Mean Square Error (MSE) yang paling kecil. Berdasarkan hasil penelitian, diperoleh kesimpulan bahwa MAF mampu meningkatkan akurasi data hingga mencapai 99.12%.Kata kunci: sensor BMP180, koreksi kesalahan, regresi linear, moving average filter, mean square error ABSTRACTThe main aspect that distinguishes sensors from one another is the level of accuracy. In this study, a system was developed to reduce the level of deviation to minimize the measurement error on low-cost sensors. The sensor used is the air pressure sensor, BMP180. Then this sensor is used to measure water depth based on air pressure in water. The Moving Average Filter (MAF) method is used to get rid of outliers of data, to obtain more relevant data for curve fitting. Then a linear regression analysis is performed to produce a function as a correction of recorded data from the sensor. System testing is carried out through a number of scenarios and then the equation is chosen with the smallest Mean Square Error (MSE). Based on this research, MAF increases data accuracy up to 99.12%.Keywords: sensor BMP180, error correction, linear regression, moving average filter, mean square error

High speed data transmision demands broader bandwidth. This has an effect towards the limitation of frequencies spectrum allocation as well as interference. To solve this, multi carrier modulation is one of choices. Having better power spectral density compared to OFDM, the FBMC-OQAM has been chosen as the multi carrier modulation. The FBMC is equipped with Poly Phase Network filter, makes it able to achieve better PSD. Frequently used filter, which is pre-emphasis, is popular in speech processing that is possibly able to be extended in use. Pre-emphasis is also accompanied with de-emphasis filter, which is similar to FBMC. Pre-emphasis filter suppresses low frequency magnitudes and emphasizes higher frequency. By assuming that noise presents in higher frequency, an approachment to protect audio signal by itself is proposed. Random noise is broadband signal where frequencies can have ranges from lower to higher with smaller magnitudes compared to signal. By providing slots in audio higher frequency and magnitudes, random noise occupied those slots with relatively weak magnitudes then sent through the air. At the receiver, a deemphasis filter invert the process to restore signal by deemphasizing higher frequency and removing noise as well. The result shows better BER with this approachment. For example, when Eb/No was 13 dB, BER with and without pre-emphasis are approximately 0.0184 dB and 0.0187 dB, consecutively. It means there was 250 bits or approximately 32 points has been corrected. It shows that pre-emphasis can work along with PPN FBMC filter to gain better BER values. Intisari-Transmisi data berkecepatan tinggi membutuhkan bandwidth yang lebih lebar. Hal ini berdampak pada terbatasnya alokasi spektrum frekuensi dan interferens. Solusinya adalah dengan menggunakan modulasi multi-carrier. Modulasi FBMC-OQAM dipilih sebagai modulasi multi-carrier yang memiliki power spectral density (PSD) lebih baik daripada OFDM. Modulasi ini menggunakan filter poly phase network sehingga PSD lebih baik. Filter pre-emphasis sering digunakan dalam pengolahan sinyal wicara dan diperluas penggunaannya dalam makalah ini.Filter pre-emphasis dilengkapi dengan filter de-emphasis yang identik dengan skema FBMC. Filter ini menekan magnitude pada frekuensi rendah dan memperkuat frekuensi tinggi. Dengan asumsi derau berada pada frekuensi tinggi, sebuah pendekatan untuk melindungi sinyal audio oleh dirinya sendiri ditawarkan. Derau acak (random noise) adalah sinyal broadband yang frekuensinya mulai dari terendah hingga tertinggi, tetapi magnitudenya lebih rendah daripada sinyal audio. Dengan menyediakan ruang spektrum pada frekuensi tinggi sinyal audio yang memiliki magnitude tinggi setelah proses pre-emphasis, derau akan masuk, tetapi magnitudenya rendah, lalu dikirimkan. Pada sisi penerima, filter de-emphasis akan mengembalikan sinyal audio dengan menekan frekuensi tinggi, yang berarti juga menekan derau. Hasil simulasi menunjukkan nilai BER yang semakin baik. Sebagai contoh, untuk Eb/No 13 dB, nilai BER dengan dan tanpa pre-emphasis...

Underwater acoustic communication is a technology that uses sound or acoustic waves and water as its propagation medium. This technology has been used in various fields, such as underwater wireless sensor networks, underwater monitoring system, and surveillance systems. An acoustic modem is required to facilitate communication between nodes. In this paper, an underwater acoustic modem using Frequency Shift Keying (FSK) modulation has been designed. This modulation is widely used because of its reliability and simple design. FSK modem was designed using M=2 level or known as Binary FSK (BFSK) with 40 kHz mark frequency and 43 kHz space frequency. This study tested data packets sending and its error rate against the distance variation. Testing for 70-bit data resulted in 1% error at 100 cm distance and 37% error at 170 cm distance. When compared with the previous testing at 1200 bps which resulted in 0% and 35% error, it can be seen that error at 1200 bps is better than at 2400 bps, but the data transmission was better at 2400 bps. Addition to the number of bits sent and distance has an influence on the error value, i.e. the greater the distance and the amount of data sent, the greater the error value.