Traffic flow monitoring is an essential component in intelligent transportation systems. However, large-scale monitoring in urban network is still a challenging task due to insufficient data. To address this challenge, this study proposes combined use of two emerging data sources: small satellites and connected vehicles (CVs). Small satellites' remote sensing provides spatial distribution of all the vehicles in wide area with certain time interval, such as a few hours. CVs' mobile sensing provides time-contentious trajectories of randomly sampled vehicles, such as several percents of penetration rate. Although each of them in itself is insufficient to estimate traffic state due to long time interval or low penetration rate, combination of these two data could be used to estimate time-continuous traffic state in wide area by complementing limitations of each other. Following this idea, a novel network traffic state estimation method is formulated. The notable feature of the proposed method is that it does not require any roadside detectors or calibrated fundamental diagram parameters, making the method applicable to any surface roads without costly sensor infrastructure. The accuracy of the proposed method was verified by conducting numerical simulation, and promising results were obtained.