With　the　emergence　of　Mobile　Edge　Computing　(MEC)　and　Ultra-Dense　Network　(UDN),　the　requirements　of　high　user　data　rate,　ultra-low　latency　and　massive　number　of　connection　in　5G　can　be　satisfied.　However,　the　access　node　selection　in　UDN　architecture　is　getting　more　complicated　due　to　the　introduction　of　computing　resources　at　the　same　time.　In　this　paper,　we　propose　an　access　node　selection　algorithm,　called　QoEANS,　that　takes　the　Quality　of　Experience　(QoE)　into　account　for　the　UDN　scenarios　with　MEC.　We　create　the　QoE　function　unified　by　Mean　Opinion　Score　(MOS)　as　the　performance　metric.　The　latency　and　energy　consumption　based　pseudo-subjective　evaluation　method　is　used　to　define　the　QoE　function　that　can　meet　the　real-time　requirements　of　access　node　selection　while　focusing　on　user　experience.　The　simulated　annealing　algorithm　is　adopted　to　solve　the　problem.　Simulation　results　show　that　the　proposed　QoEANS　algorithm　can　improve　the　QoE　significantly,　and　perform　better　than　the　basic　access　node　selection　strategy,　Maximum　Signal　to　Interference　plus　Noise　Ratio　(MaxSINR).　©　2018　IEEE.