The L1 legacy Satellite-Based Augmentation System (SBAS) does a good job of augmenting the Global Positioning System (GPS). However, the L1 SBAS Minimum Operational Performance Standards (MOPS) can fully support only GPS, and since some parameters (e.g. PRN Mask) for supporting multiple constellations are not defined in L1 MOPS, the advantages of multi-constellation cannot be utilized. To complement this disadvantage, research on the Dual-Frequency, Multi-Constellation (DFMC) SBAS, the 2nd generation SBAS, is currently ongoing. According to the GNSS User Technology Report published by European Space Agency (ESA), only about 15% of commercial GNSS receivers support both L1 and L5 signals. Also, it is expected that the L1 legacy SBAS systems will continue to provide L1 SBAS service even after the introduction of DFMC service to maintain backwards compatibility with L1 SBAS users. Therefore, it is necessary to improve the performance of the L1 legacy SBAS for L1 SBAS users as well as DFMC users. Single-Frequency, Multi-Constellation (SFMC) Satellite-Based Augmentation System (SBAS) proposed in 2017 (Lim, 2017) is a method that enables the L1 legacy SBAS to provide multi-constellation GNSS augmentation service. Two keys of the SFMC SBAS concept are 1) removing MT 2-5 which currently occupy more than 50% of SBAS message bandwidth in order to broadcast MT 25 for satellite error corrections of other GNSS constellations and 2) modifying MT 6 to provide user differential range error (UDRE) for multi-GNSS. The role of providing pseudo-range correction (PRC) and UDRE should be transferred to MT 25 and MT 6 due to the removal of MT 2-5. Although legacy SBAS separates fast clock correction (PRC) of MT 2-5 from long-term clock correction term (?af0) of MT 25, it is possible to include the PRC into clock correction of MT 25 in SA-off era because both parameters are inherently a correction for satellite clock error. MT 6 should be used to update the UDRE of multi-constellation GNSS instead of MT 2-5. Since Issue Of Data Fast corrections (IODF) in the current MT 6 is no longer necessary due to removal of MT 2-5, the IODF can be replaced with Issue Of Data PRN mask (IDOP) for correcting the UDRE to PRN mask of MT 1. In this study, we conducted experiments to analyze the performance improvement of legacy SBAS by the SFMC concept using actual legacy SBAS messages and user measurements. We used Wide Area Augmentation System (WAAS) in US, European Geostationary Navigation Overlay System (EGNOS) in Europe, Multi-functional Satellite Augmentation System (MSAS) in Japan as the legacy SBAS systems for the test, and three reference stations (SGPO for WAAS, MEDI for EGNOS, DAEJ for MSAS) capable of receiving each SBAS geostationary satellite signals were chosen as SBAS users. The user observation data were obtained from the receiver independent exchange (RINEX) observation data files logged at each reference station for 24 hours from 00:00:00 to 23:59:59 in March 10, 2019, and SBAS broadcast data (PRN 135 for WAAS, PRN 123 for EGNOS, PRN 129 for MSAS) were from the CNES Navigation and Time Monitoring Facility (NTMF) FTP server. In addition, we generated satellite ephemeris and clock error corrections for multi-GNSS (e.g. Galileo, BDS) to be included in MT 25 using precise orbit and clock products from Wuhan University GNSS Research Center (GRC) FTP server. For three cases of constellation combination (case 1: GPS, case 2: GPS+Galileo, case 3: GPS+Galileo+BDS), we analyzed the performance of SFMC SBAS and legacy SBAS in terms of accuracy, integrity, availability, message bandwidth. Comparing 95th percentile error, the performance of SFMC SBAS was better than that of legacy SBAS in all three SBAS systems. In particular, 95th percentile errors of MSAS for case 1, case 2, case 3 were respectively 1.15m, 0.95m, 0.87m horizontally and 1.32m, 1.25m, 1.26m vertically. Moreover, the maximum position error of the legacy SBAS user has been reduced by 30~40% after applying SFMC to the current SBAS, which means the SFMC contributes to provide accurate position more robustly. According to Stanford plot analysis for the integrity performance (CAT-I for WAAS and EGNOS, LPV200 for MSAS), the ratio of nominal operations of the SFMC SBAS was far higher than the legacy SBAS. In case of WAAS user, SPGO, the ratio of nominal operations for case 1, case 2, case 3 was 63.217, 99.965%, >99.999%, and the number of samples that satisfy CAT-I integrity requirement has significantly increased after the application of SFMC SBAS concept. For an availability analysis, we computed the ratio of areas with an availability above 99% in each SBAS service area, and the results showed that the SFMC SBAS contribute to expand the areas that meet the availability requirement. Even though there is currently no airport that can provide CAT-I service in U.S., however the ratio of CAT-I service-available area to the contiguous United States (CONUS) can be enlarged to 98.53%, 100% by adding Galileo and Galileo+Beidou respectively. Lastly, analysis results of the bandwidth occupation per message type confirms that it is possible to add augment message for Galileo and BDS constellations in 250 bps using empty time slots saved by the removal of MT 2-5. In the case of EGNOS and MSAS, extra one or more constellation beside GPS, Galileo and Beidou can be added.