Low-cost Indoor location positioning has been explored to provide an exact location inside a house or a building where GPS signal cannot be obtained. Among many technologies, wireless range-based method and Dead Reckoning (DR) utilizing accelerometer are widely investigated. However, there are several problems in wireless range-based method. First, the position estimation accuracy is low because it must need environment equipped with multiple Access Point (AP). Second, accurate position estimation is difficult because of buildings and walls which causes multi-path fading; sometimes, wireless signal does not reach a measuring device. In contrast, DR is not dependent on locations and provides almost the same accuracy inside a building. Since relative distance is calculated utilizing an accelerometer. However, high-accuracy location estimation is difficult because cumulative errors are generated by accelerometer. Based on the above background, we first report results of preliminary investigations about differences of cumulative errors generated by DR between smartphones. Then, we propose a cumulative error compensation model based on preliminary investigations and evaluate the experimental results.